There is a whole giant essay I probably need to write at some point, but I can't help but see parallels between today and the Industrial Revolution.
Prior to the industrial revolution, the natural world was nearly infinitely abundant. We simply weren't efficient enough to fully exploit it. That meant that it was fine for things like property and the commons to be poorly defined. If all of us can go hunting in the woods and yet there is still game to be found, then there's no compelling reason to define and litigate who "owns" those woods.
But with the help of machines, a small number of people were able to completely deplete parts of the earth. We had to invent giant legal systems in order to determine who has the right to do that and who doesn't.
We are truly in the Information Age now, and I suspect a similar thing will play out for the digital realm. We have copyright and intellecual property law already, of course, but those were designed presuming a human might try to profit from the intellectual labor of others. With AI, we're in the industrial era of the digital world. Now a single corporation can train an AI using someone's copyrighted work and in return profit off the knowledge over and over again at industrial scale.
This completely unpends the tenuous balance between creators and consumers. Why would a writer put an article online if ChatGPT will slurp it up and regurgitate it back to users without anyone ever even finding the original article? Who will contribute to the digital common when rapacious AI companies are constantly harvesting it? Why would anyone plant seeds on someone else's farm?
It really feels like we're in the soot-covered child-coal-miner Dickensian London era of the Information Revolution and shit is gonna get real rocky before our social and legal institutions catch up.
> Prior to the industrial revolution, the natural world was nearly infinitely abundant. We simply weren't efficient enough to fully exploit it.
This is just wildly incorrect. People started running out of trees during the early Iron Age. Woodlands have been a managed and often over exploited resource for a long time. Active agriculture vs passive woodlands vs animal grazing has been in constant tension for thousands of years across most of the globe.
I believe running out of trees was always a local issue - there weren't enough trees where you were at because getting trees had to be gotten locally, you didn't go get trees from far away. So yes that was in constant tension, the thing is that the problem of having enough trees turned from a local problem to a global problem, with the side effects of not having enough trees globally that the world needed to maintain the environment humanity first conquered.
I think the natural world was nearly infinitely abundant is a reasonable description, resource depletion was always local before mass industrialization. Being able to exploit the world as opposed to just your local area is also a mark of efficiency.
The general point is accurate, don’t take it so literally.
There were more than enough trees until we developed the technology to clear cut in expeditious manner. There were more than enough fish until we developed the technology to pull massive indiscriminate amounts out of the ocean (and/or started polluting our rivers with industry). There was more than enough topsoil until we developed mechanized plows and artificial fertilizer. Etc.
A few hundred years ago or less, a squirrel could get from the Atlantic Ocean to the Mississippi River without ever touching the ground. Not possible today. That’s not a push and pull played out over thousands of years, that’s a one-way trend.
This is where STEM people are weak- a lack of knowledge on history. In another forum, someone would have chipped in that England's virgin forests were fully deforested by 1150. And someone else would have pointed out that this deforestation produced the economic demand for coal that drove the Industrial Revolution in the first place.
Still, that kind of underscores OP's point. Yes, natural resources were not completely unlimited prior to the Industrial Revolution; Jonathan Swift predated Watt's steam engine, after all. Still... Neither were information resources 10 years ago. Intellectual property laws did exist prior to AI, of course. The legal systems in place are not completely ignorant of the reality.
However, there's an immense difference in scale between post-industrial strip mining of resources, and preindustrial resource extraction powered solely by human muscle (and not coal or nitrogylcerin etc). Similarly, there's a massive difference in information extraction enabled by AI, vs a person in 1980 poring over the microfilm in their local library.
The legal system and social systems in place prior to the Industrial Revolution proved unsuitable for an industrial world. It stands to reason that the legal system and social systems in today's society would be forced to evolve when exposed to the technological shift caused by AI.
Most of humanity survived on agriculture and sometimes hunting-gathering for last 10k years. People that survived on hunting whales is minuscule. Comparing those two is nonsensical.
The industrial revolution is generally understood to have started somewhere around 1760, Moby Dick took place in approximately 1830, about 10 years before what some historians mark as the end of the agrarian to Industrial shift that is generally termed the Industrial revolution
I get sort of wishy-washy from 1830 on, because lots of people put the end of the Industrial revolution as being 1900, but 1840 is a defensible and commonly held position.
from an global perspective it isn't. Some places sure, like Western Europe, who in some cases had completed enclosure, but remember the new world had only been discovered a few hundred years ago at that point.
Just google maps the north part of South America, even today there are large swathes of undeveloped land across it and back then it was considerably less exploited. At that time it would have appeared infinite, especially to the European industrialists.
we're talking about the fucking industrial revolution, of course this defaults to the European perspective. Unless you wanna spit some new bars about Aztec foundries and train lines connecting meso-america in the 19th century, then the point stands. At that time, the world appeared to the industrialists of the industrial revolution to be infinite. Nor had humanity discovered the terrible side effects of fossil fuels on the atmosphere.
Sure, of course it's convenient to ignore the native peoples and pretend that prior to the Industrial Revolution the rest of the world outside of Europe was some untapped well of resources that Europeans had a natural right to.
Who might be swept underfoot in this "Information Revolution", I wonder?
Yes, just the other day I saw someone make a comment about write performance in SQLite without considering the plight of the Baltic peoples in the Northern Crusades. It was really convenient of them to do that, fucking typical.
How do you think they're enabling the mass surveillance tech? SQLite got reach bruv.
Your continued erasure of the Baltic people's continues to cut deep into my heart, and your callous candour to their plight, as you discard any chance to mention them, continues to shock me.
Nobody said anything about Europeans having a "natural right". Bad enough to derail a conversation with irrelevant political nitpicking, unforgiveable to use a strawman to do so. Boo.
GP made a comparison between what we're going through and the Industrial Revolution. Ignoring the negatives of that revolution - like by acting as though the "new world" was uninhabited/unused and so Europeans had a right to its resources - seems like a bad idea.
"but I can't help but see parallels between today and the Industrial Revolution"
You're not the only one.
The current Pope Leo XIV explicitly named himself after the the previous Leo, Pope Leo XIII, who was pope during the Industrial Revolution (1878-1903) and issued the influential Encyclical Rerum novarum (Rights and Duties of Capital and Labor) in response to the upheaval.
“Pope Leo XIII, with the historic Encyclical Rerum novarum, addressed the social question in the context of the first great industrial revolution,” Pope Leo recalled. “Today, the Church offers to all her treasure of social teaching in response to another industrial revolution and the developments of artificial intelligence.” A name, then, not only rooted in tradition, but one that looks firmly ahead to the challenges of a rapidly changing world and the perennial call to protect those most vulnerable within it.”
As you know, I deeply respect you. Not trying to argue here, just provide my own perspective:
> Why would a writer put an article online if ChatGPT will slurp it up and regurgitate it back to users without anyone ever even finding the original article?
I write things for two main reasons: I feel like I have to. I need to create things. On some level, I would write stuff down even if nobody reads it (and I do do that already, with private things.) But secondly, to get my ideas out there and try to change the world. To improve our collective understanding of things.
A lot of people read things, it changes their life, and their life is better. They may not even remember where they read these things. They don't produce citations all of the time. That's totally fine, and normal. I don't see LLMs as being any different. If I write an article about making code better, and ChatGPT trains on it, and someone, somewhere, needs help, and ChatGPT helps them? Win, as far as I'm concerned. Even if I never know that it's happened. I already do not hear from every single person who reads my writing.
I don't mean that thinks that everyone has to share my perspective. It's just my own.
Agreed, totally! I still write and put stuff online.
But it definitely feels different now. It used to feel like I was tending a public garden filled with other people who might enjoy it. It still kind of feels like that, but there are a handful of giant combine machines grinding their way around the garden harvesting stuff and making billionaires richer at the same time.
It's not enough to dissuade me from contributing to the public sphere, but the vibe is definitely different.
Honestly, it reminds me a lot about the early days of Amazon. It's hard to remember how optimistic the world felt back then, but I remember a time when writing reviews felt like a public good because you were helping other people find good products. It was like we all wanted honest product information and Amazon provided a neutral venue for us to build it. Like Wikipedia for stuff.
But as Amazon got bigger and bigger and the externalities more apparent, it felt less like we were helping each other and more like we were help Bezos buy yet another yacht or media empire. And as the reviews got more and more gamed by shady companies, they became less of a useful public good. The whole commons collapsed.
I worry that the larger web and digital knowledge environment is going that way.
I still intend to create and share my stuff with the world because that's who I want to be. But I'll always miss the early days of the web where it felt like a healthier environment to be that kind of person in.
> But as Amazon got bigger and bigger and the externalities more apparent, it felt less like we were helping each other and more like we were help Bezos buy yet another yacht or media empire.
The Internet-circulating quote comes to mind: Planet Earth is pretty much a vacation resort for around 500 rich people, and the remaining 8 billion of us are just their staff. The Relative Few have got the system set up perfectly so that whatever we do, we're probably serving/enriching them. AI doesn't really change this, but it does further it.
> The Internet-circulating quote comes to mind: Planet Earth is pretty much a vacation resort for around 500 rich people, and the remaining 8 billion of us are just their staff. The Relative Few have got the system set up perfectly so that whatever we do, we're probably serving/enriching them. AI doesn't really change this, but it does further it.
I don't necessarily disagree with the analysis on how Planet Earth is currently setup to be, but something that I've been thinking about lately, is that to the extent we can consume the public image of some of the Relative Few, they seem oddly unhappy.
Anyone who finds themselves with $100m in their bank account and thinks, "No, I need more," is a person with a hole inside them that can never be filled.
There were a couple of threads on HN this week. "Do you have any unusual hobbies" and "How do you relax". I enjoyed them and was thinking of contributing. Then it occurred to me that my comments would be gold for targeting advertising at me. That is the distrust that has been bred by the data harvesters.
It does feel like the collaborative, free open nature of the web has gone and the optimism that brought… it feels like no one would build Foursquare today.
But then I wonder if I’m just old an jaded and to the younger generation creating content, for them the web is open and expressive- just in a different way
I can totally see that, for sure. I was much more likely to write a review long ago, now I don't even bother. (For buying stuff online, at least.) Maybe I lost my innocence about this stuff a long time ago, and so it's not so much LLMs that broke it for me, but maybe... I dunno, the downfall of Web 2.0 and the death of RSS? I do think that the old internet, for some definition of "old," felt different. For sure. I'll have to chew on this. I certainly felt some shock on the IP questions when all of this came up. I'm from the "information wants to be free" sort of persuasion, and now that largely makes me feel kinda old.
Also I'm not a fan of billionaires, obviously, but I think that given I've worked on open source and tools for so long, I kinda had to accept that stuff I make was going to be used towards ends I didn't approve of. Something about that is in here too, I think.
(Also, I didn't say this in the first comment, but I'm gonna be thinking about the industrial revolution thing a lot, I think you're on to something there. Scale meaningfully changes things.)
I feel the future includes the sentiments you describe. It was a little before my time professionally, but I grew up reading that kind of thinking.
I do think that the open web stuff, decentralized, or at least more decentralized than currently, is the path forward. I've been reading about the AT protocol and it recently becoming an official working group with the IETF.
I feel a second order effect of making decentralized social networking easier, is making individuals more empowered to separate from what they don't believe in. The third order effect is then building separate infrastructure entirely.
As sad as that can be - in my personal opinion it runs the risk of ending the "world wide" part of the web - it appears to be the only way society can avoid enriching the few beyond reason.
> I'm from the "information wants to be free" sort of persuasion, and now that largely makes me feel kinda old.
Me too, 100%. But that was during a moment in time when that information was more likely to be enabling a person who otherwise didn't have as many resources than enabling a billionaire to make their torment nexus 0.1% more powerful.
> I kinda had to accept that stuff I make was going to be used towards ends I didn't approve of. Something about that is in here too, I think.
Yeah, I've mostly made peace with that too.
The way I think about it is that when I make some digital thing and share it with the world, I'm (hopefully!) adding value to a bunch of people. I'm happiest if the distribution of that value lifts up people on the bottom end more than people on the top. I think inequality is one of the biggest problems in the world today and I aspire to have the web and the stuff I make chip away at it.
If my stuff ends up helping the rich and poor equally and doesn't really effect inequality one way or the other, I guess it's fine.
But in a world with AI, I worry that anything I put out there increases inequality and that gives me the heebie-jeebies. Maybe that's just the way things are now and I have to accept it.
> But in a world with AI, I worry that anything I put out there increases inequality and that gives me the heebie-jeebies. Maybe that's just the way things are now and I have to accept it.
This observation doesn't really clash with "information wants to be free." You just have to include LLMs in the category or "information," like Free Software types already do for all software. You don't need to abandon your principles, you should shift your demands. A handful of companies can't be allowed to benefit from free information and then put what they make behind a wall.
Free Software types also create software...they didn't just argue for a better license and try to regulate Sun/others to re-license their software; they wrote free (libre) versions of proprietary software and released it for free (cost), which is what counteracted the "[putting] what they make behind a wall". If you're saying "[some] LLMs should be free", I agree.
I don’t disagree with you, but this has been going on for a while…
Google monetized the the by indexing it and monetized what you wanted to find.
Facebook monetized the eyeballs from the pictures and posts you added.
Now LLMs will monetize all web content.
To play devil’s advocate - LLMs do give something back. Those with ideas and no coding experience can now build entire businesses for little to zero cost. This seems different
> the "information wants to be free" sort of persuasion
That was always a luxury of its peculiar historical moment, though, wasn't it? Barlow didn't have to care who paid for the infrastructure, but he was just bloviating.
No, it's as true now as it was then. The intellectual property team didn't win on the merits or by law enforcement; it was the convenience of streaming anything at will for a monthly fee that did the trick.
> it was the convenience of streaming anything at will for a monthly fee that did the trick
That's not the whole story, though. There have been many community-driven projects to bring convenient access to copyrighted works to the masses in a convenient way. You may recall the meteoric success of Popcorn Time. Law enforcement shut them down. Without the hand of the state beating down any popular alternative to legal distribution it absolutely would be the dominant mode of media consumption.
> It used to feel like I was tending a public garden filled with other people who might enjoy it. It still kind of feels like that, but there are a handful of giant combine machines grinding their way around the garden harvesting stuff and making billionaires richer at the same time.
An underrated upside to being harvested is that your voice has now effectively voted in the formation of the machine's constitution. In a broader ecological sense, you've still tended to a public garden, but in this case your work is part of the nutrient base for a different thing.
Broader still: after the machines squeeze all of our inputs into an opaque crystal, that crystal's very purpose is to leak it all back out in measured doses. Yes, "some billionaire" will own the lion's share of that process, but time so far is telling that efforts can be made to distill strong, open, public versions of the same.
If raw resources (tree cutting) and manufacturing (book binding) is saturated, a fully-realized economy has just one step left: financialization.
You have to start finding ways to keep people hooked on books and make it a part of their regular lifestyle. One book can't be enough, and after a while you have to convince them to replace the books they already bought. New editions, Author's Footnotes, limited run release, all of the stops have to be pulled out to get consumers to show up en-masse. Because that's what they are - consumers, not readers - wallets to be squeezed until they're bled of all the trust they had in media.
I think about the publications I liked reading as a kid, like Joystiq and Polygon. Some of the best games journalism the industry produced, but inevitably doomed to fail as their competitors monetized further. The rest of traditional media has followed the same path, converging on some mercurial social network marketing tactic as the placeholder for big-picture brand strategy.
> A lot of people read things, it changes their life, and their life is better. They may not even remember where they read these things. They don't produce citations all of the time. That's totally fine, and normal. I don't see LLMs as being any different. If I write an article about making code better, and ChatGPT trains on it, and someone, somewhere, needs help, and ChatGPT helps them? Win, as far as I'm concerned. Even if I never know that it's happened. I already do not hear from every single person who reads my writing.
Not a contradiction but an addendum: plenty of creative pursuits are not about functional value, or at least not primarily. If somebody writes a seemingly genuine blog post about their family trauma, and I as the reader find out it's made-up bullshit, that's abhorrent to me, whether or not AI is involved. And I think it would be perfectly fair for writers who do create similar but genuine content to find it abhorrent that they must compete with genAI, that genAI will slurp up their words, and that genAI's mere existence casts doubt on their own authenticity. It's not about money or social utility, it's about human connection.
The consent question gets weirder when agents have persistent memory. I run agents that accumulate context over weeks — beliefs extracted from observations, relationships with other agents. At what point does an agent's
memory become its own work product vs. derivative of its training? There's no legal framework for that.
Consent is absolutely important, but that does not mean that every single thing in the entire world requires explicit consent. You did not ask me for consent to use my words in your comment. That does not mean you're a bad person.
Free use is an important part of intellectual property law. If it did not exist, the powerful could, for example, stifle public criticism by declaring that they do not consent to you using their words or likeness. The ability to do that is important for society. It is also just generally important for creating works inspired by others, which is virtually every work. There has to be lines for cases where requiring attribution is required, and cases where it is not.
> You did not ask me for consent to use my words in your comment.
I am not representing your words as mine. I am not using your words to profit off. I am not making a gain by attributing your words to you.
> There has to be lines for cases where requiring attribution is required, and cases where it is not.
You are blurring the lines between "using a quote or likeness" and "giving credit to". I am skeptical that you don't know the difference between the two.
Regardless, any "perspective" that disregards the need to acquire consent is invalid. Even if you are going to ignore it, you have to acknowledge that you don't feel you need any consent from the people you are taking from.
This whole "silence is consent" attitude is baffling.
You made an incredibly strong statement that is much broader than what we are talking about. I am pointing out various cases where I think that broadness is incorrect, I am not equating the two.
I do not think that, if you read, say, https://steveklabnik.com/writing/when-should-i-use-string-vs... , and then later, a friend asks you "hey, should I use String or &str here?" that you need my consent to go "at the start, just use String" instead of "at the start, just use String, like Steve Klabnik says in https://steveklabnik.com/writing/when-should-i-use-string-vs... ". And if they say "hey that's a great idea, thank you" I don't think you're a bad person if you say "you're welcome" without "you should really be saying welcome to Steve Klabnik."
It is of course nice if you happen to do so, but I think framing it as a consent issue is the wrong way to think about it.
We recognize that this is different than simply publishing the exact contents of the blog post on your blog and calling it yours, because it is! To me, an LLM is a transformative derivative work, not an exact copy. Because my words are not in there, they are not being copied.
But again, I am not telling anyone else that they must agree with me. Simply stating my own relationship with my own creative output.
he doesn't have solid points, he conflates fair use with free use (?), ignores thousands of years of attribution history, and equates normal human to human learning with corporate LLMs training on original content (without consent). Great presentation, like you said, to cover the logical defects.
I did say "free use" instead of "fair use," yeah. That's my mistake, thank you for the correction. If I could edit my original comment, I would, mea culpa. Typos happen.
Fair use of training data hasn’t yet been settled in court. People here are treating it like it has been. But no amount of wishful thinking or moral arguments will change a verdict saying it’s fine for training data to be used as it has been.
Until that question is settled, it’s disingenuous to dismiss his points out of hand as conflating fair use or ignoring consent.
However, I don't feel comfortable suggesting that this is settled just yet, one district judge's opinion does not mean that other future cases may disagree, or we may at some point get explicit legislation one way or the other.
I think the court dropped the ball here. On the one hand, I think they were right that using existing works--copyrighted or otherwise--to train a model was transformable fair use. On the other hand, Anthropic and others trained their models on illicit copies of the works; they (more often than not) didn't pay the copyright holders.
There's a doctrine in Fifth Amendment law called "fruit of the poisonous tree." The general rule is that prosecutors don't get to present evidence in a criminal trial that they gained unlawfully. It's excluded. The jury never gets to see it even if it provides incontrovertible evidence of guilt. The point is to discourage law enforcement from violating the rights of the accused during the investigative process, and to obtain a warrant as the Amendment requires.
It seems to me that the same logic ought to be applied to these companies. They want to make money by building the best models they can. That's fine! They should be able to use all the source data they can legitimately obtain to feed their training process. But if they refuse to do so and resort to piracy, they mustn't be allowed to claim that they then used it fairly in the transformative process.
I was just enumerating some of the issues with the '''solid''' points OP made. Actually addressing them would take too long and be exercise in futility, here, in HN, in april 2026. Why would I put in the effort, for my comment to be flagged and sent to the void? or worse, persisted for ever and used for training without my consent?
And yes, you are right, the legal and moral question of fair use in training data hasn't been settled yet; we agree here.
I don't think the poster has a viewpoint that 'refuses consent', their viewpoint is their writing they put for others to view is for others to view, regardless of how it is viewed. They seem to be giving consent, not refusing it, no?
Prior to Industrial Revolution, nobody could go hunt in the woods, because the woods were King’s, and poaching King’s game carried death penalty. Situation was similar on the continent: the tiny slivers of remaining wood lands were off limits.
Granted, things were different in the New World, as a result of mass depopulation event following the Columbian exchange. But even there, the megafauna was hunted to extinction soon after the humans first appeared there.
Anyway, the point is that no, prior to Industrial Revolution, the world was of full of scarcity, not abundance.
I'm halfway through Foundation on Apple TV and this piece landed hard (you had me at Asimov) because of it. Asimov's whole deal with psychohistory is that you can predict what large populations do even when individuals are unpredictable. Seldon doesn't need anyone to be honest; he needs the math to converge on something real about how people actually behave.
LLMs are sort of the inverse of that. They produce text that looks like the statistical aggregate of human knowledge, but nothing underneath is converging on truth. Seldon's math worked because it modeled actual dynamics. LLMs work because they model plausible text. The "jagged competence frontier" Kingsbury describes: crushing multivariable calculus, failing a word problem, is exactly what you'd get from a system that learned the shape of correct answers without learning what makes them correct.
The part of Foundation that feels prescient right now isn't the predicting-the-future stuff. It's the part where everyone can see Empire is hollowing out and the response is to just...keep going. More spectacle, more confidence, less substance holding any of it up. Hmmm, wait...
Can we interpret "abundant" in a Darwinian sense e.g. diversity of life? I would think the industrial farming revolution decreased crop variety over time same for animal lineages aside from the rapid increase in mixed poodle breeds.
To add, I don’t think my ancestor Spaniards for example needed the help of machines to deplete mines in America. They also came already equipped with all kinds of legal systems, including the Requerimiento, which they read out loud to natives in preposterous spectacle.
In general the transition from feudalism to capitalism, including the formation of the legal systems that supported the latter, happened gradually for maybe up to four or five centuries before the steam engine had been invented.
Sure, the Industrial Revolution further accelerated the development of property rights, mercantile, and civil laws, but all in all I don’t think there’s much truth that machines were the primary cause of such developments.
Not really Malthusian. Agricultural societies had adapted to keep the population stable during normal times and bounce back in a generation or two after bad times. Those cultural adaptations stopped working when childhood mortality declined.
Useful land was a scarce resource in more civilized regions, while labor was cheap. Given enough land, subsistence farmers could easily feed themselves outside particularly bad years. But much of the land belonged to local elites, and commoners had to work that land to fund the pursuits of the elites.
If I'm being honest, I've never related to that notion of remuneration and credit being the primary reason to write something. I don't claim to be some great writer or anything, but I do have a blog I write quite often on (though I'm traveling in my wife's Taiwan now and haven't updated it in a while). But for me, I write because it feels good to do so. Sometimes there's a group utility in things like I edit a Google Maps listing to be correct even though "a faceless corporation is going to hoover up my work and profit off it without paying me for my work" and I might pick up a Lime bike someone's dropped into the sidewalk even though "a faceless corporation is externalizing the work of organizing the proper storage of their property on public land without paying the workers" or so on.
I just think it's nice to contribute to the human commons and it's fine if some subset of my fellow organism uses it in whatever way. Realistically, the fact that Brewster Kahle is paid whatever few hundred thousand he's paid for managing a non-profit that only exists because it aggregates other people's work isn't a problem for me. Or that Larry Page and Sergey Brin became ultra-rich around providing a search interface into other people's work. Or that Sam Altman and Dario Amodei did the same through a different interface.
This particular notion doesn't seem to be a post-AI trend. It seems to have happened prior to the big GPTs coming out where people started doing a lot of this accounting for contribution stuff. One day it'll be interesting to read why it started happening because I don't recall it from the past. Perhaps I just wasn't super plugged in to the communities that were complaining about Red Hat, Inc.
It's not that I don't understand if I sold my Subaru to a guy who immediately managed to sell it to another guy for a million times the money. I get that. I'd feel cheated. But if I contributed a little to it, like I did so Google would have a site to list for certain keywords so that they could show ads next to it in their search results, I just find it so hard to be like "That's my money you're using. Pay me!".
You do it as a hobby, that's fine. Some people do it for a living. And while they aren't owed a living doing that specific thing, it is going to be a big problem for them if they can't make money at it anymore.
I'm sure plenty of people feel the same way about software. They make software as a hobby and don't care about remuneration or credit. Meanwhile I write software for my day job and losing the ability to make money from it would be devastating.
I was going to write, "not for long," which might be true for some. But then I realized there will always be a difference between LLM output and human writing. We don't read blogs because of their facts, we read them because of how the facts are presented and how the author's personality comes through on the page.
EDIT: That said, LLMs are great at faking it, and a lot of amateur writing will be difficult to distinguish from LLM output. So I'm disagreeing with myself a bit.
But we are talking about "slurping up" IP and regurgitating it right? OK. So if I slurp up Mickey Mouse and output Micky Mouse that's an offense. But what if I slurp up a billion images and output some chimera? That's what the LLMs do. And that's what humans do too.
Ah, I see. It’s just straightforward protectionism like dockworkers opposing automation and so on. That I do comprehend, in fact.
I write software too and I may no longer be able to just do it in the old way. Pretty scary world but also exciting. I can’t imagine trying to restrict LLM software writers on that basis but I can comprehend it as simply self-interest.
Do you make money writing software? I bet you either try to restrict LLM usage or assign your rights to an employer who does. Putting code in the public domain is pretty rare, and extremely rare for paid work.
If your code isn't in the public domain, then anything you haven't explicitly allowed them to train on is restricted for them. They've been ignoring that for anything they can actually get their hands on, but it's there.
Maybe it’s a useless distinction but it feels like we’ve gone through overlapping ages of Communication, Data, Processing, and Information.
First we conquered the ability to move matter and transmit signal, greatly shrinking the world. Next was sensor technology, especially the mobilization of it, and our ability to collect more data than we could ever imagine being able to process. Then we started going crazy with data centres and big data and the idea that maybe we can somehow correlate it all if we just process it enough. And now we’re finally turning data into information, building enormous graphs of correlation without even having to manually reason about a lot of it. Before AI, the hard part was figuring out how to go about finding the signal you needed. Now it’s getting easier at an incredible speed.
At what point do we look at 'Industrial Society and its Future' and go from "yeah that'll never happen", "ok some parts of it are happening", to ...? I swear tech folks are the most obtuse people on the planet.
I think it's completely normal. Whenever automation comes knocking, people are inclined to think it's going to flatline conveniently before their job is at risk. LLMs can code now? Cool, they can't code well though can they? Oh they can code pretty well now? Cool, coding was never the hard part of SWE anyway, it's [thing we have no reason to think AI can't beat 99% of humans at at some point], etc
I think SWE as a mainstream profession is much nearer to the end than the beginning, I'm curious and quite scared about what becomes of us.
I don't think you understand. Frankly, AI is a failure if all it does is replace coders. AI needs (given its current investment levels) to conquer all forms of knowledge work. This is an example of tech/industry needing to impose itself on society, rather than society needing it.
That's how human progress works. No one can want or need it because they cannot conceptualize wanting it until someone shows that it is possible. Now, many of those wants become needs.
We can absolutely conceptualize what we want or need. I was born in 1980 in NYC. When I was a boy my father took me to a tech conference where they had a demo of ordering TV shows on demand. It was a miracle, to my young mind. Was this what I needed?
Growing up I had a friend group of misfit boys, who discovered h4ck1ng and phr34king. But we also discovered slackware Linux on 3.5" floppies. We also had to discover ASM and compiling the linux kernel in order to do anything with it. Boys with machines. That wasn't what I needed either.
Later on we did have great things with tech. Google made the world searchable in ways Altavista didn't. I remember strapping the original iPod on my arm to go for runs outside. I didn't even need a car for a while investors subsidized my Uber rides to and from the office.
Now, it seems the US is balanced on a precipice. The economy seems to have an incredible amount of money desperate to grow, but to what purpose. In my lifetime, and in my parents, and their parents before them, when the dollar becomes restless the flag goes forth. The dollar follows the flag.
You wouldn't have known about a TV had you not seen it. That is what I mean by, people generally can't conceptualize what they want or need until they see it.
My point was not about the difference, it was about the fact that average people cannot conceptualize new ideas until one person or team invents it, then the average person will want or need it.
As for AI, I and many others want it, and some even need it, in certain use cases. Speak for yourself.
I believe the idea that you (or I) might know better than the 'average people' to be incredibly conceited, arrogant, and frankly wrong. It is an attitude that gives you superiority for having achieved nothing.
I think your numbers are off. TAM for office workers is ~20T a year, of which SWE compensation is ~3T. So if they can make 3T x 10% X 5 years = 1.5T that covers their current valuations. It's not as insane as you make out, even not taking into account the other high risk areas like legal, accounting etc
Hit the nail on the head with that framing. So many articles are now coming out addressing the anxieties about adoption of a new technology, but we genuinely don’t really need it as a society.
I still wonder if we really needed the iPhone or many other things we’re told is “progress” and innovation in an arrow of time manner. The future is not set in stone and things need not play out in this manner at all. Unlike the iPhone where most were excited by its possibilities (even if they traded precious privacy in the name of convenience), there’s not a clear reason that this version of LLM driven technologies represent significant upsides than downsides.
>This completely unpends the tenuous balance between creators and consumers. Why would a writer put an article online if ChatGPT will slurp it up and regurgitate it back to users without anyone ever even finding the original article? Who will contribute to the digital common when rapacious AI companies are constantly harvesting it? Why would anyone plant seeds on someone else's farm?
I have been thinking about this. I was pretty amendment a few months ago that AI is going to make a lot of thing worse for everyone because of the externalities of the technology (Data Center Creep, lock in of models, ect) and it probably still will. But then someone suggested to me that I use Claude Code to upgrade my SSG site to the new version because I had been sitting on my ass as the years went by, missing deadline by deadline. I just couldn't put my self into gear to upgrade it. It was massively out of date 10 years plus and I knew it was going to be a nightmare to deal with the problems. I probably was making it more harder than it really was in my head.
So I purchase Claude Code pro and the thing upgrade my site pretty well. There were things it missed because I didn't know the problems existed in the first place until the upgrade was complete, but I had a working updated site in less than an hour. If I had done this myself it would have taking me days/weeks.
So at that point I realized something. Its a tool that can handle good amount of tasks I throw at it as long as I am specific. I think the problem with most people is they expect it to respond like a human. Thats not going to happen, IMHO. Maybe some day it will be more than what it is but right now its just a tool. I don't care what anyone says about AGI and the likes. Its not going to happen with the current iteration (the pattern recognition type) We are going to need more than that if we want to simulate a human brain..
The point is. And I know this is not going to be received very well, mostly because this tech is in the hands of people that are gatekeeping it, is that maybe someday we might reach a point where all of humanities knowledge is put into these things and we can use them to better our lives. Maybe at some point we don't need to hold onto or hoard things as if its the only way we can make a living? And instead we can build things just for the sake of creating it and improve humanity in the process? Obviously the commercial model of these things is not great, that is going to have to be dealt with, but I can see a future where we might be able to fix a lot of humanities problems with this technology as more and more good people put it to use for things that help humanity.
> If all of us can go hunting in the woods and yet there is still game to be found, then there's no compelling reason to define and litigate who "owns" those woods.
Property rights don't just protect natural resources, but labor as well. If I cleared out hunting ground in that forest to be the prime spot to catch animals, I would make sure I can use it when I want.
> a small number of people were able to completely deplete parts of the earth
A small number of people seems inaccurate when there's typically many more individuals in the pipelines for these technologies.
> and in return profit off the knowledge over and over again at industrial scale
Not off just that knowledge, there needed to be a model trained on the data of many others to utilize it.
> Why would a writer put an article online if ChatGPT will slurp it up and regurgitate it back to users without anyone ever even finding the original article?
Who's better at writing in this scenario and what are my motivations? If it's ChatGPT and I did it for money, then I would say I should recognize that I can't compete and find something AI can't do. If it's ChatGPT and I write to convey my ideas in an effort to learn regardless of the bestowment of a new perspective on the reader, I'll keep writing.
> Why would anyone plant seeds on someone else's farm?
They wouldn't unless it was their own way to attain food and survive. And if it's not the only way, they can defer to those with optimal methods to get it the cheapest they can in the market.
Mostly, AIs don’t recite back various works. Yes, there a couple of high profile cases where people were able to get an AI to regurgitate pieces of New York Times articles and Harry Potter books, but mostly not. Mostly, it is as if the AI is your friend who read a book and gives you a paraphrase, possibly using a couple sentences verbatim. In other words, it probably falls under a fair use rule.
Secondly, given the modern world, content that doesn’t appear online isn’t consumed much, so creators who are doing it for the money will certainly continue putting content online. Much of that content will be generated by AIs, however.
You're missing the point. This is the crux of munificent's argument IMO (and I've made variations of it as well)
> We have copyright and intellecual property law already, of course, but those were designed presuming a human might try to profit from the intellectual labor of others.
You getting a summary of a copyrighted work from a friend is necessarily limited by the number of friends you have, the amount of time they have to read stuff and talk to you, and so on. Machines (and AIs) don't have any limitations.
Yes, true. But does that really shift the argument much? An AI is like the most well-read book nerd you’ve ever met. The AI has read everything. They still won’t recite Harry Potter for you at full length and reading what the original author wrote is part of the pleasure.
Sure, we could change current law, but I think that only forces an AI company to buy one copy of every book. I don’t think it gives any sort of royalty stream to anyone beyond that. Copyright is literally the right to make copies. Once I have acquired a copy, I can read it, summarize it, transform it, etc. in myriad ways.
Does a literal book nerd profit megacorporations when they bring up books to you? While burning through a household worth of energy in the process?
Also, I’d like to talk with such book nerd because they’d have opinions on books, potentially if I brought up something I have read we could exchange thoughts about it, they could make recommendations for me based on their complex experiences instead of statistics from Reddit comments. An LLM can do none of those, while also doing the former. It’s a lose-lose.
Also, a book nerd doesn’t take roughly ~all human created text to train to produce meaningful results. It’s just such a misplaced analogy and people have been making it ever since OpenAI announced chatgpt for the first time - why do people think “an LLM is just a human who read a lot”
> Megacorporations making profit is not some evil that needs to be stopped.
Externalities are a thing. It's not about the profit per se, but about how (a) the making of that profit might negatively impact others, and (b) the deployment of that profit in pursuit of rent-seeking and other antisocial behavior in order to insure its continued existence might also negatively impact others.
> If all of us can go hunting in the woods and yet there is still game to be found, then there's no compelling reason to define and litigate who "owns" those woods.
> Why would a writer put an article online if ChatGPT will slurp it up and regurgitate it back to users without anyone ever even finding the original article?
In the brave new world we're creating, people will write specifically for AI. If you can impress models so much that they "regurgitate" your work, then your work has achieved a kind of immortality.
> We have copyright and intellecual property law already, of course, but those were designed presuming a human might try to profit from the intellectual labor of others. With AI, we're in the industrial era of the digital world. Now a single corporation can train an AI using someone's copyrighted work and in return profit off the knowledge over and over again at industrial scale.
The idea that copyright simply doesn't apply to AI has more to do with AI companies deciding that they're not going to comply with those laws than the design of the laws. Also a very successful lobby against enforcement by positioning AI as a strategic necessity.
> Prior to the industrial revolution, the natural world was nearly infinitely abundant. We simply weren't efficient enough to fully exploit it. That meant that it was fine for things like property and the commons to be poorly defined. If all of us can go hunting in the woods and yet there is still game to be found, then there's no compelling reason to define and litigate who "owns" those woods.
I mean, medieval Europe (speaking broadly) had pretty well defined property rights wrt hunting. In fact, the forester at the time was thought of as one of the most corrupt jobs, as they'd commonly have side hustles poaching and otherwise illegally extracting resources from the lands they enforced and kept others from utilizing in a similar way. Quis custodiet ipsos custodes?
The natural world was not meaningfully abundant… Way before the industrial revolution land which was once used for opening hunting was closed off by the ruling class. Even before the Industrial Revolution you had a new class of merchant and factory owners who earned riches to buy land and keep the poor from hunting on it. Much of the natural resources out of reach for the majority and only accessible by those with deep pockets
> Why would a writer put an article online if ChatGPT will slurp it up and regurgitate it back to users without anyone ever even finding the original article?
I'm happy to miss all the stuff that was written just for the financial benefit of the author.
> We are truly in the Information Age now, and I suspect a similar thing will play out for the digital realm.
The analogy seems to be backwards though. It would be as if we previously had a scarcity of land and because of that divided it up into private property so markets could maximize crop yield etc. and then someone came up with a way to grow food on asteroids using robots, and that food is only at the 20th percentile of quality but it's far cheaper. Suddenly food becomes much more abundant and the people who had been selling the 20th percentile food for $5 are completely out of the market because the new thing can do that for $0.05, and the people providing the 50th percentile food for $10 are also taking a hit because the price difference between what they're providing and the 20th percentile stuff just doubled.
The existing plantation owners then want to put a stop to this somehow, or find a way to tax it, but arguments like this have a problem:
> Why would a writer put an article online if ChatGPT will slurp it up and regurgitate it back to users without anyone ever even finding the original article?
This was already the status quo as a result of the internet. Newspapers were slowly dying for 20 years before there was ever a ChatGPT, because they had been predicated on the scarcity of printing presses. If you published a story in 1975 it would take 24 hours for relevant competitors to have it in their printed publication and in the meantime it was your exclusive. The customer who wants it today gets it from you. On top of that, there weren't that many competitors covering local news, because how many local outlets are there with a printing press?
Then blogs, Facebook, Reddit and Twitter come and anyone who can set up WordPress can report the news five minutes after you do -- or five hours before, because now everyone has an internet-connected camera in their pocket so the first news of something happening now comes in seconds from whoever happened to be there at the time instead of the next morning after a media company sent a reporter there to cover it.
The biggest problem we have yet to solve from this is how to trust reports from randos. The local paper had a reputation to uphold that you now can't rely on when the first reports are expected to come from people with no previous history of reporting because it's just whoever was there. But that's the same thing AI can't do either -- it's a notorious confabulist.
And it's the media outlets shooting themselves in the foot with this one, because too many of them have gotten far too sloppy in the race to be first or pander to partisans that they're eroding the one advantage they would have been able to keep. Damn fools to erode the public's trust in their ability to get the facts right when it's the one thing people would otherwise still have to get from them in particular.
I struggle so hard with this anthropomorphism of LLMs. At the end of the day it's a statistical gradient descent predictor with a bunch of "shit" bolted on top to try and steer outputs in a specific way.
They don't have the actual concept of "benevolent"... or a concept of anything at all. Based on an input, they regress down a path of "what is the next most probable statistical token to output next" and that's fucking it, with the bolted-on shit manipulating these outputs a bit.
I don't doubt that at some point there will be some other AI leap, but I'm not even sure it'll be built on this foundation.
What really needs to be developed is an actual artificial brain of sorts. Much like an infant learns language from first principals, a real AI would have a phase of continuous growth, creating actual memories and being able to reflect upon them. I daresay context windows are not that.
I'd really like to encourage anyone to pump the brakes a bit on how these things actually work, and what they actually are. There is a reason sama is pivoting away from video, et. al. and into corporate software coding, much like anthropic.
Stuff gets put online when the reader isn't the customer. Someone is paying for a reader to be told certain things. So it's free at the point of reading.
> It really feels like we're in the soot-covered child-coal-miner Dickensian London era of the Information Revolution and shit is gonna get real rocky before our social and legal institutions catch up
The really discouraging part of this is that it feels like our social and legal institutions don't even care if they catch up or not.
Technology is speeding up and the lag time before anything is discussed from a legal standpoint is way, way too long
I've been looking at things from the same lens since 2023. At the same time, the depletion/hoarding bit isn't new. Companies were already doing this with consumer data, LLMs are just finally the factory moment—now that we have all the raw material we finally have a means of automating production using it.
So, in some ways, I also view LLMs as a pivotal and important wake up call. Companies were already taking the data and using it for a variety of other purposes—it was just way less evident to people when they weren't in direct competition with labor, since, under capital, labor is what we sell.
Either an entire new industry needs to form, or it's finally time to move beyond capitalism. Centralized capital ends up killing itself, because it effectively shuts down its own engine if it kills off consumers, who can only exist in the first place if the wage labor structure holds.
> 2017’s Attention is All You Need was groundbreaking and paved the way for ChatGPT et al. Since then ML researchers have been trying to come up with new architectures, and companies have thrown gazillions of dollars at smart people to play around and see if they can make a better kind of model. However, these more sophisticated architectures don’t seem to perform as well as Throwing More Parameters At The Problem. Perhaps this is a variant of the Bitter Lesson.
This is not true and unfortunately this significantly reduced the credibility of this article for me. Raw parameter counts stopped increasing almost 5 years ago, and modern models rely on sophisticated architectures like mixture-of-experts, multi-head latent attention, hybrid Mamba/Gated linear attention layers, sparse attention for long context lengths, etc. Training is also vastly more sophisticated.
The Bitter Lesson is misunderstood. It doesn't say "algorithms are pointless, just throw more compute at the problem", it says that general algorithms that scale with more compute are better than algorithms that try to directly encode human understanding. It says nothing about spending time optimising algorithms to scale better for the same compute, and attention algorithms and LLMs in general have significantly advanced beyond "moar parameters" since the time of Attention is All You Need/GPT2/GPT3.
Literally the paragraph right before the one you quote is this:
> I am generally outside the ML field, but I do talk with people in the field. One of the things they tell me is that we don’t really know why transformer models have been so successful, or how to make them better. This is my summary of discussions-over-drinks; take it with many grains of salt. I am certain that People in The Comments will drop a gazillion papers to tell you why this is wrong.
As I understand it, this article is basically a conglomeration of several attempts at an article that the author has attempted to make over the past decade or so considering the impacts of AI on society. In their own words:
> Some of these ideas felt prescient in the 2010s and are now obvious. Others may be more novel, or not yet widely-heard. Some predictions will pan out, but others are wild speculation. I hope that regardless of your background or feelings on the current generation of ML systems, you find something interesting to think about.
As for the "Bitter Lesson" part, they pretty much directly said that it wasn't the Bitter Lesson exactly, saying it might be a variant of it. Honestly, it felt more like a way of throwing in a reference to something that also might provoke thought, which was done throughout the piece (which again, is the entire point).
It's totally valid to say "this article didn't provoke much thought for me". I'm a bit confused at why you think a lack of specific domain knowledge in a domain that they literally state they are not an expert in would be disqualifying for that purpose though.
The title of the article is “The Future of Everything is Lies, I Guess” and the first part is literally complaining about LLMs being bullshit machines, while the author proceeds to tell confabulations (or lies) of his own. Is there not a bit of irony in that?
If you’re a non-expert in a field, I don’t think it’s a good sign if you’re writing a 10 part article about that field’s impact on society and getting basic facts wrong. How can I trust that the conclusions will be any more credible?
Might the conclusions be correct even if some of the facts are not? Even a stopped clock is right twice a day. And, "approximately correct" is still sometimes valuable.
>Raw parameter counts stopped increasing almost 5 years ago, and modern models rely on sophisticated architectures like mixture-of-experts, multi-head latent attention, hybrid Mamba/Gated linear attention layers, sparse attention for long context lengths, etc.
Agree, I recently updated our office's little AI server to use Qwen 3.5 instead of Qwen 3 and the capability has considerably increased, even though the new model has fewer parameters (32b => 27b)
Yesterday I spent some time investigating it:
- Gated DeltaNet (invented in 2024 I think) in Qwen3.5 saves memory for the KV kache so we can afford larger quants
- larger quants => more accurate
- I updated the inference engine to have TurboQuant's KV rotations (2026) => 8-bit KV cache is more accurate
Before, Qwen3 on this humble infra could not properly function in OpenCode at all (wrong tool calls, generally dumb, small context), now Qwen 3.5 can solve 90% problems I throw at it.
All that thanks to algorithmic/architectural innovations while actually decreasing the parameter count.
5 years ago was the beginning of 2021, just under a year after GPT3 was released (which was not good at doing anything useful). And that model was 175B params.
GPT4 has been widely rumored to have 1.8 trillion params, which is 10x more, and was released 2 years after this "5 years ago" date that you are using here.
So, to quote yourself here, "This is not true and unfortunately this significantly reduced the credibility of this article for me" /s/article/comment
In late 2021, GLaM had 1.2T parameters. It's difficult to find much use of it in the wild and while the benchmarks it uses are rather outdated, it has a HellaSwag score of 76.6% and WinoGrande of 73.5%. GPT3 had 64.3% and 70.2%.
Meanwhile, Gemma 2 9B, a model from July 2024 with 133x fewer parameters than GLaM, scores 82% and 80.6%. Hellaswag and WinoGrande aren't used in modern benchmarks, probably because they're too easy and largely memorised at this point.
And GPT-4 had 1.8T parameters sure, but it's noticeably worse than any modern model a fraction of the size, and the original incarnation was ridiculously expensive per token. And in any case, its number of parameters was only possible due using mixture-of-experts, which I would definitely classify as a sophisticated architecture as opposed to just throwing more parameters at a vanilla transformer. Even in 2021 GLaM was a MoE because the limits of scaling dense transformers had already been hit.
MoE has made it vastly easier to increase total parameters (and recent open models are really quite large) but it's also hard to compare a MoE with an earlier dense model.
Yeah I also came here to be one of those People In The Comments the author refers to.
Transformers are not magical. They are just a huge improvement over other architectures at the time such as LSTMs and RNNs and even CNNs. They allowed us to throw more and more compute at the problem of next token prediction. And we’ve been riding that horse ever since.
Another big advancement that deserves mentioning is “reasoning” models that have the opportunity to spit out thinking tokens before giving a final answer.
None of this is to say transformers are the most principled approach. But they work.
Transformers' greatest improvement over RNN/LSTM was to enable better parallelization of large-scale training. This is what enabled language models to become "large". But when controlling for overall size, more RNN/LSTM-like approaches seem to be more efficient, as seen e.g. in state space models. The transformer architecture does add some notable capabilities in accounting for long-range dependencies and "needle in a haystack" scenarios, but these are not a silver bullet; they matter in very specific circumstances.
With modern training techniques, RNNs (not just linear SSMs, potentially even vanilla LSTMs) can scale just as well as transformers or even better when it comes to enormous context lengths. Dot-product attention has better performance in a number of domains however (especially for exact retrieval) so the best architectures are likely to remain hybrid for now.
>With modern training techniques, RNNs (not just linear SSMs, potentially even vanilla LSTMs) can scale just as well as transformers or even better when it comes to enormous context lengths.
That's not true. Modern training techniques aren't enough. Vanilla RNNs with modern training techniques still scale poorly. You have to make some pretty big architectural divergences (throwing away recurrency during training) to get a RNN to scale well. None of the big labs seem to be bothered with hybrid approaches.
> That's not true. Modern training techniques aren't enough. Vanilla RNNs with modern training techniques still scale poorly. You have to make some pretty big architectural divergences (throwing away recurrency during training) to get a RNN to scale well.
SSMs move the non-linearity outside of the recurrence which enables parallelisation during training. It is trivial to do this architectural change with an LSTM (see the xLSTM paper). Linear RNNs are still RNNs.
But you can still keep the non linearity by training with parallel Newtown methods, which work on vanilla LSTMs and scale to billion of parameters.
> None of the big labs seem to be bothered with hybrid approaches.
Does Alibaba not count? Qwen3.5 models are the top performers in terms of small models as far as my tests and online benchmarks go.
>SSMs move the non-linearity outside of the recurrence which enables parallelisation during training. It is trivial to do this architectural change with an LSTM (see the xLSTM paper). Linear RNNs are still RNNs.
Removing the non-linearity from the recurrence path is exactly what constitutes a "pretty big architectural divergence." A linear RNN is an RNN in a structural sense, certainly, but functionally it strips out the non-linear state transitions that made traditional LSTMs so expressive, entirely to enable associative scans. The inductive bias is fundamentally altered. Calling that simply 'modern training techniques' is disingenous at best.
>But you can still keep the non linearity by training with parallel Newtown methods, which work on vanilla LSTMs and scale to billion of parameters.
That does not scale anywhere near as well as Transformers in compute spend. It's paper/research novelty. Nobody will be doing this for production.
>Does Alibaba not count? Qwen3.5 models are the top performers in terms of small models as far as my tests and online benchmarks go.
I guess there's some misunderstanding here because Qwen is 100% a transformer, not a hybrid RNN/LSTM whatever.
> That does not scale anywhere near as well as Transformers in compute spend. It's paper/research novelty. Nobody will be doing this for production.
What exactly makes you so confident?
The world is not just labs that can afford billion dollar datacentres and selling access to SOTA LLMs at $30/Mtokens. Transformers are highly unsuitable for many applications for a variety of reasons and non-linear RNNs trained via parallel methods are an extremely attractive value proposition and will likely feature in production in the next products I work on.
> I guess there's some misunderstanding here because Qwen is 100% a transformer, not a hybrid RNN/LSTM whatever.
See the Qwen3.5 Huggingface description: https://huggingface.co/Qwen/Qwen3.5-27B
> Efficient Hybrid Architecture: Gated Delta Networks combined with sparse Mixture-of-Experts deliver high-throughput inference with minimal latency and cost overhead.
> It remains unclear whether continuing to throw vast quantities of silicon and ever-bigger corpuses at the current generation of models will lead to human-equivalent capabilities. Massive increases in training costs and parameter count seem to be yielding diminishing returns. Or maybe this effect is illusory. Mysteries!
I’m not even sure whether this is possible. The current corpus used for training includes virtually all known material. If we make it illegal for these companies to use copyrighted content without remuneration, either the task gets very expensive, indeed, or the corpus shrinks. We can certainly make the models larger, with more and more parameters, subject only to silicon’s ability to give us more transistors for RAM density and GPU parallelism. But it honestly feels like, without another “Attention is All You Need” level breakthrough, we’re starting to see the end of the runway.
I see a lot of researchers working on newer ideas so I wouldn't be surprised if we get a breakthrough in 5-10 years. After all, the gap between AlexNet and Attention is All You Need was only 6 years. And then Scaling Laws was about 3-4 years after that. It might seem like not much progress is being made but I think that's in part because AI labs are extremely secretive now when ideas are worth billions (and in the right hands, potentially more).
Of course 5-10 years is a long time to bang our heads against the wall with untenable costs but I don't know if we can solve our way out of that problem.
I think in domains like Math and Software Engineering, they are less constrained by training data anyway. They can synthetically generate and validate programs. To what extent that scales into novel insights is a different matter, but I think they dream of the AlphaGo Zero moment at least in verifiable domains.
Based on what's happened so far, maybe. At least that's exactly how we got to the current iteration back in 2022/2023, quite literally "lets see what happens when we throw an enormous amount data at them while training" worked out up until one point, then post-training seems to have taken over where labs currently differ.
Right, but we played the scaling card and it worked but is now reaching limits. What is the next card? You can surely argue that we can find a new one at any time. That’s the definition of a breakthrough. I just don’t see one at the moment.
Did you see the one before the current one was even found? Things tend to look easy in hindsight, and borderline impossible trying to look forward. Otherwise it sounds like you're in the same spot as before :)
That’s what I’m said. Breakthroughs happen. No doubt about it, and they are unpredictable. Hence a breakthrough. But right now we’re using up runway with nothing yet identified to take us to the next level. And while sometimes breakthroughs happen, sometimes they don’t.
We pay people to create more high quality tokens (mercor, turing) which are then fed into data generating processes (synthetic data) to create even more tokens to train on
But does that really help, or do you get distortion? The frequency distribution of human generated content moves slowly over time as new subjects are discussed. What frequency distribution do those “data generating processes” use? And at root, aren’t those “data generating processes” basically just another LLM (I.e., generating tokens according to a probability distribution)? Thus, aren’t we just sort of feeding AI slop into the next training run and humoring ourselves by renaming the slop as “synthetic data?” Not trying to be argumentative. I’m far from being an AI expert, so maybe I’m missing it. Feel free to explain why I’m wrong.
That's the problem in a nutshell. There is an art to how you generate the synthdata so that you don't get crappy trained models (especially when mistakes cost XX million dollars).
It's also theoretically why facebook paid 14bn for alex wang and scale ai
> The current corpus used for training includes virtually all known material.
This is just totally incorrect. It's one of those things everyone just assumes, but there's an immense amount of known material that isn't even digitized, much less in the hands of tech companies.
The amount of private data that is locked up inside private internal databases is huge. This is especially true of regulated industries. There is a wealth of data - financial data showing how to budget for things, pricing data on various products that are B2B, standard operating procedures at mature companies that have gone through various revisions, designs for manufacturing plants so people don't keep reinventing and making the same mistakes again, and on and on.
Reddit alone contains about the same quantity of text (~10 billion posts * 10 words per post, vs 1 million books * 100k words per book). Messaging and document platforms (google docs, slack, discord, telegram, etc.) probably each have 1-3 orders of magnitude more than reddit. To your/GP's point though, those private platforms probably haven't been slurped up by LLMs yet.
Which is what percent of the world’s content? 0.000000001% or something similar. It’s nothing in the scheme of things. To put it another way, if we were to digitize that continent and train on it, our AIs would not get noticeably better in any way. It doesn’t move the needle.
You’re missing the point. And we’re not just talking about books, whatever that might mean. We’re talking about all documents ever made. Every magazine article, every blog and web page, every Word doc, etc. I’m pretty sure that whatever is in the Vatican archives is tiny by comparison. Given the age of the Vatican archives, I can also guarantee that many of those “books” are nothing more than page fragments. Very few will be full codices or long scrolls. Many will date before the printing press when document production was slow and laborious.
I think the discussion has to be more nuanced than this. "LLMs still can't do X so it's an idiot" is a bad line of thought. LLMs with harnesses are clearly capable of engaging with logical problems that only need text. LLMs are not there yet with images, but we are improving with UI and access to tools like figma. LLMs are clearly unable to propose new, creative solutions for problems it has never seen before.
I'm not sure if you misread the statement you quoted or I'm misreading yours, but it doesn't sound like you're really disagreeing with their point. Did you miss the "un" in "unable", or am I misunderstanding you as also saying that you don't consider them to be creative?
> LLMs with harnesses are clearly capable of engaging with logical problems that only need text.
To some extent. It's not clear where specifically the boundaries are, but it seems to fail to approach problems in ways that aren't embedded in the training set. I certainly would not put money on it solving an arbitrary logical problem.
> To some extent. It's not clear where specifically the boundaries are, but it seems to fail to approach problems in ways that aren't embedded in the training set. I certainly would not put money on it solving an arbitrary logical problem.
In what way can you falsify this without having the LLM be omniscient? We have examples of it solving things that are not in the training set - it found vulnerabilities in 25 year old BSD code that was unspotted by humans. It was not a trivial one either.
Here's an odd example of testing, but I design very complex board and card games, and LLMs are terrible at figuring out whether they make sense or really even restating the rules in a different wording.
I thought they would be ideal for the job, until I realized that it would just pretend that the rules worked because they looked like board game rules. The more you ask it to restate, manipulate or simulate the rules, the more you can tell that it's bluffing. It literally thinks every complicated set of rules works perfectly.
> it found vulnerabilities in 25 year old BSD code that was unspotted by humans.
I don't think the age of the code makes the problem more complex. Finding buffers that are too small is not rocket science, bothering to look at some corner of some codebase that you've never paid attention to or seen a problem with is. AI being infinitely useful (cheap) to sic on pieces of codebase nobody ever carefully looks at is a great thing. It's not genius on the part of the AI.
> Here's an odd example of testing, but I design very complex board and card games, and LLMs are terrible at figuring out whether they make sense or really even restating the rules in a different wording.
I'm positive that they are perfectly fine and will a pretty good job. Did you actually try it?
Eh, I can see their point, I think. The models can restate the rules differently, I'm sure, but it sounds like the GP is saying that LLMs can't tell whether the rules are well-balanced.
It would be interesting to see some example problems along those lines. Design some games with complex rules, including one or two of the most subtle game-wrecking bugs you can think of, and ask the models if they can spot them.
In fact that sounds more interesting the more I think about it. Intensive RL on that sort of thing might generalize in... let's say useful ways.
Let’s be careful. That’s a straw man. I don’t know anyone who says that. Aphyr says in the article that AIs can do things. But they have been marketed as “intelligent,” and I agree with Aphyr that the word is suggesting way more than AIs currently deliver. They do not reason and they do not think and are not truly intelligent. As the article says, they are big wads of linear algebra. Sometimes, that’s useful.
We know that they do not reason because we know the algorithm behind the curtain. The model is generating the next token via model weights and some randomness. That’s all. It not reasoning. Sometimes it has an appearance of reasoning, but not if you know how it works. It doesn’t matter that the model manufacturer marketing department slaps a “Reasoning!” sticker on the side of the model. It’s not actually doing that. As an analogy, sometimes a stage magician in Las Vegas makes it seem that he’s making a woman disappear and a tiger appear in her place, but we all know that’s not what is really happening; It’s just a clever trick.
Well, could you define what reasoning actually means? What would an AI need to do to be considered capable of reasoning? What is the core difference between what we do that is considered reasoning verse what AI currently does that is not considered reasoning?
To be clear, I am not making a statement as to whether AI reasons or not. Its just slippery to say something isn't or can't do X when we can't really define X. Perhaps if we can put it down as an outcome rather than an, in my opinion, currently impossible to accurately define characteristic of a thing.
I guess it's a form of engagement to give a wildly wrong answer, but I'm not convinced that the extra nuance you've introduced is really all that nuanced either.
I keep explaining to my peers, friends and family that what actually is happening inside an LLM has nothing to do with conscience or agency
and that the term AI is just completely overloaded right now.
One thing that has happened is that "AI" has been an academic discipline since literally the 1950s. The term was originally used in the hope that we would soon be able to emulate human minds. This turned out to be hard, but the name stuck to the discipline.
Now, suddenly, this name has been broadcast to every human in the world more or less. To them, it's a new term, and it obviously means something human mind-like. But to people who work on AI, that's not generally what it means. (Which isn't to say that some of them don't think we're near to achieving that; they just use other terms like "AGI" for that goal). So the name, which has a long history, is deceptive to people who aren't familiar with computer science.
> Now, suddenly, this name has been broadcast to every human in the world more or less. To them, it's a new term, and it obviously means something human mind-like. But to people who work on AI, that's not generally what it means. (Which isn't to say that some of them don't think we're near to achieving that; they just use other terms like "AGI" for that goal). So the name, which has a long history, is deceptive to people who aren't familiar with computer science.
I think it's even worse than that: people were familiar with the term already, but from science fiction, where it referred to actually human-level intelligence. It's similar to the "hoverboard" thing from a while back, except this time with profoundly higher stakes and requires for more technical knowledge to be able to see that it is in fact touching the ground.
AI is exactly the right term: the machines can do "intelligence", and they do so artificially.
Just like we have machines that can do "math", and they do so artificially.
Or "logic", and they do so artificially.
I assume we'll drop the "artificial" part in my lifetime, since there's nothing truly artificial about it (just like math and logic), since it's really just mechanical.
No one cares that transistors can do math or logic, and it shouldn't bother people that transistors can predict next tokens either.
> Just like we have machines that can do "math", and they do so artificially.
Nobody calls calculators "artificial mathemeticians", though; we refer to them by a unique word that defines what they can and can't do in a far less fanciful and ambiguous way.
> AI is exactly the right term: the machines can do "intelligence", and they do so artificially.
AI in pop culture doesn't mean that at all. Most people impression to AI pre-LLM craze was some form of media based on Asmiov laws of robotics. Now, that LLMs have taken over the world, they can define AI as anything they want.
In 2018, ie “pre-LLM”, the label “AI” was already stamped to everything, so I highly doubt that most people thought that their washing machines are sentient in any way. I remember this starkly, because my team was responsible at Ericsson (that time, about 120k employees) for one of the crucial step to have models in production, and basically every single project wanted that stamp.
The shift in meaning has been slowly diluted more and more across decades.
Structurally a transformer model is so unrelated to the shape of the brain there's no reason to think they'd have many similarities. It's also pretty well established that the brain doesn't do anything resembling wholesale SGD (which to spell it is evidence that it doesn't learn in the same way).
>Structurally a transformer model is so unrelated to the shape of the brain there's no reason to think they'd have many similarities.
Substrate dissimilarities will mask computational similarities. Attention surfaces affinities between nearby tokens; dendrites strengthen and weaken connections to surrounding neurons according to correlations in firing rates. Not all that dissimilar.
Given the large visible differences in behavior and construction, akin to the difference between a horse and a pickup truck, I would ask the reverse question: In what ways do LLMs meet the definition of having consciousness and agency?
Veering into the realm of conjecture and opinion, I tend to think a 1:1 computer simulation of human cognition is possible, and transformers being computationally universal are thus theoretically capable of running that workload. That being said, that's a bit like looking at a bird in flight and imagining going to the moon: only tangentially related to engineering reality.
> In what ways do LLMs meet the definition of having consciousness and agency?
Agency: an ability to make decisions and act independently. Agentic pipelines are doing this.
Consciousness: something something feedback[1] (or a non-transferable feeling of being conscious, but that is useless for the discussion). Recurrent Processing Theory: A computation is conscious if it involves high-level processed representations being fed back into the low-level processors that generate it.
Tokens are being fed back into the transformer.
> that's a bit like looking at a bird in flight and imagining going to the moon: only tangentially related to engineering reality.
Is it? Vacuum of space is a tangible problem for aerodynamics-based propulsion. Which analogous thing do we have with ML? The scaled-up monkey brain[2] might not qualify as the moon.
If platonic representation hypothesis holds across substrates, then it might matter very little, in the end. It holds across architectures in ML, empirically.
The crowd of "backpropagation and Hebbian learning + predictive coding are two facets of the very same gradient descent" also has a surprisingly good track record so far.
For starters, natural brains have the innate ability to differentiate between things that it knows and things that it have no possibility of knowing...
Does it matter? Evolution is the brain's very own "pre-training". Hundreds of millions of years of priors hardwired.
We can do that for AIs too - pre-train on pure low Kolmogorov complexity synthetics. The AI then "knows things" before it sees any real data. Advantageous sometimes. Hard to pick compute efficient synthetics though.
> Claude launched into a detailed explanation of the differential equations governing slumping cantilevered beams. It completely failed to recognize that the snow was entirely supported by the roof, not hanging out over space. No physicist would make this mistake, but LLMs do this sort of thing all the time.
You have to meet some physicist friends of mine then. They are likely to assume that the roof is spherical and frictionless.
I think it's too early to declare the Turing test passed. You just need to have a conversation long enough to exhaust the context window. Less than that, since response quality degrades long before you hit hard window limits. Even with compaction.
Neuroplasticity is hard to simulate in a few hundred thousand tokens.
I think for a while the test was passed. Then we learned the hallmark characteristics of these models, and now most of us can easily differentiate. That said -- these models are programmed specifically to be more helpful, more articulate, more friendly, and more verbose than people, so that may not be a fair expectation. Even so, I think if you took all of that away, you'd be able to differentiate the two, it just might take longer.
Right. I think the modern LLMs are quite good at mimicking human words, but we were initially taken in like we were in the 1960s by ELIZA. It’s a (increasingly sophisticated) magic trick, but it’s just a trick.
For as rigorous of a Turing test as you present, I believe many (or even most) humans would also fail it.
How many humans seriously have the attention span to have a million "token" conversation with someone else and get every detail perfect without misremembering a single thing?
Response quality degrades long before you hit a million tokens.
But sure, let's say it doesn't. If you interact with someone day after day, you'll eventually hit a million tokens. Add some audio or images and you will exhaust the context much much faster.
However, I'll grant you that Turing's original imitation game (text only, human typist, five minutes) is probably pretty close, and that's impressive enough to call intelligence (of a sort). Though modern LLMs tend to manifest obvious dead giveaways like "you're absolutely right!"
It's weird, I don't know how normally pedantic comp sci. people let this meme that the Turing test is beaten by LLMs to spread so unchallenged. As far as I'm aware, there is no restriction in the Turing test that demands that the interrogator be ignorant of the latest state-of-art in computing (and AI tech), nor is there a strict time limit enforced for the questioning?
Given these conditions, it should be relatively easy for the interrogator to expose the AI in this current day and age.
There are a lot of difference kinds of LLMs. 0 of the ones I've encountered are good writers, in fact all of them are horrible at it.
But I wonder if there's one out there that I don't know about with a different kind of training that actually is good at writing and fun to talk to for a long time. (granted somepeople love talking to gpt 4, but also some people loved talking to ELIZA so clearly some people have a super high tolerance for slop.)
I don't know. Practically, LLMs are already better conversation partners on any topic compared to the average human I have access to. This also holds in reverse, of course - if someone wants me to explain something, usually they'd be better off asking an LLM.
I hesitate to tamper with an internet master's title, but "The Future of Everything is Lies, I Guess" doesn't really summarize what in fact is a balanced, informed overview which (to me at least) is above the median for one of these thought pieces. Since it's also baity and the HN guidelines ask for such titles to be rewritten, I've taken the license.
In such cases we always try to find a phrase from the article itself which expresses what it's saying in a representative way. (There nearly always is one.) In this case, both the very first and very last sentences do this, and it's interesting that they more or less agree. So I plucked the last sentence and put it above.
Edit: oof, I missed that this is actually the first part of a long series. Not sure what we'll do about the others; I expect some of those will make the frontpage as well.
The article has a good take on the "lie" problem. We know about the hallucination problem, which remains serious. The "lie" problem mentioned is that if you ask an LLM why it said or did something, it has no information of how it got a result. So it processes the "why" as a new query, and produces a plausible explanation. Since that explanation is created without reference to the internals of how the previous query was processed, it may be totally wrong.
That seems to be the type of "lie" the author is worried about in this essay.
Thank you both! I totally missed the sidebar on the OP which explains that this is Part 1 of what will be a long series. Not sure how we'll handle that...
Articles like this should approach topics on consciousness with more humility than is displayed here.
We don’t even agree on a good definition of what’s going on inside our own heads yet, what gives you the confidence to say that what goes on inside an LLM can’t be conscious?
Obviously, the LLMs lack the divine spark, so they can't be conscious. Same as clones, IVF babies, or half of all the twins.
Jest aside, I do agree. If you list out every prominent theory of consciousness, you'd find that about a quarter rules out LLMs, a quarter tentatively rules LLMs in, and what remains is "uncertain about LLMs". And, of course, we don't know which theory of consciousness is correct - or if any of them is.
With LLMs, materialists have got a silicone idol to worship. They believe that they know everything about this idol because they've created it, and at the same time they believe that LLMs have a secret sauce that makes it intelligent. Looking at how they are trying to extract intelligence from it reminds me of alchemists of the past who tried to extract gold from lead.
Even if nothing substantial come out of this,
having shortest paths to the corpus of all human expressions in all languages! and media formats is quite something by itself, what could be the ultimate hard information retrieval tool is hiding those trace behind untracked convolutions is the real shame here. Found so much real information in between less and less hallucinations already that was impossible to retrieve otherwise in that time frame.
Basically tokenrank kills pagerank.
If you change it from asking a question to giving an instruction, how many humans do you know that have trouble saying no to things that aren't reasonable? I'd argue that pretty much every human will refuse to do most things you might instruct them to do, whereas an LLM will happily attempt most things you ask them to do for you, regardless of whether they're capable of succeeding, and it's up to you to figure out if they actually did it right or not. There are tasks where this is extremely useful, but they're ones that are extremely low risk and can easily be audited upon completion. This isn't anywhere near the level of what a human is capable of.
This is a limitation of the training data. If you were uncertain about something, you wouldn’t write a book about it. The kinds of people you’re talking about tend to generate far more text in their lives than others, because they can spend more time generating - writing books, blogposts, whatever - and less time thinking and working and actually doing things. The models never say they’re uncertain because we never say we’re uncertain, or at least we don’t write it down anywhere.
Some people point at LLMs confabulating, as if this wasn’t something humans are already widely known for doing.
I consider it highly plausible that confabulation is inherent to scaling intelligence. In order to run computation on data that due to dimensionality is computationally infeasible, you will most likely need to create a lower dimensional representation and do the computation on that. Collapsing the dimensionality is going to be lossy, which means it will have gaps between what it thinks is the reality and what is.
They key capability that humans have that I've yet to see in an LLM is the ability to recognize when they would not be capable of doing a task well and refuse to do it poorly instead. The only times I've ever seen LLMs give up on a problem are when the prompting is very explicitly crafted to try to elicit a response like that when necessary or after very long back-and-forth exchanges where they get repeated feedback about unsatisfactory results. I think this has pretty dire implications in terms of what the consequences are for deploying them in any scenario where failure has significant risk or the output can't be immediately audited for correctness.
The concern for me about LLMs confabulating is not that humans don't do it. It's that the massive scale at which LLMs will inevitably be deployed makes even the smallest confabulation extremely risky.
I don't understand this. Many small errors distributed across a large deployment sounds a lot like normal mode of error prone humans / cogs / whatevers distributed over a wide deployment.
There's a difference between 1000 diverse humans with varied traits making errors that should cancel out because of the law of large numbers vs 10 AI with the same training data making errors that would likely correlate and compound upon each other.
Let's say a given B2B system deployment typically requires 100 custom behaviours/scripts and 3 years worth of effort. A team of ten people can execute such a deployment in 3-4 months. The team has the capacity to fix up issues caused by small human errors as they arise, since they show up roughly once a week.
With the advent of LLMs, a new deployment now takes 3 days. Consequently, errors requiring human attention crop up several times a day.
I have yet to see a comparison of human vs. LLM confabulation errors at scale.
"Many small errors" makes a presumption about LLM confabulation/hallucination that seems unwarranted. Pre-LLM humans (and our computers) have managed vast nuclear arsenals, bioweapons research, and ubiquitous global transport - as a few examples - without any catastrophic mistakes, so far. What can we reasonably expect as a likely worst case scenario if LLMs replacing all the relevant expertise and execution?
Your project vue-skuilder has 6 github action steps devoted to checking the work you do before it's allowed to go out. You do not trust yourself to get things right 100% of the time.
I am watching people trust LLM-based analysis and actions 100% of the time without checking.
> Some people point at LLMs confabulating, as if this wasn’t something humans are already widely known for doing.
I think we need to start rejecting anthropomorphic statements like this out of hand. They are lazy, typically wrong, and are always delivered as a dismissive defense of LLM failure modes. Anything can be anthropomorphized, and it's always problematic to do so - that's why the word exists.
This rhetorical technique always follows the form of "this LLM behavior can be analogized in terms of some human behavior, thus it follows that LLMs are human-like" which then opens the door to unbounded speculation that draws on arbitrary aspects of human nature and biology to justify technical reasoning.
In this case, you've deliberately conflated a technical term of art (LLM confabulation) with the the concept of human memory confabulation and used that as a foundation to argue that confabulation is thus inherent to intelligence. There is a lot that's wrong with this reasoning, but the most obvious is that it's a massive category error. "Confabulation" in LLMs and "confabulation" in humans have basically nothing in common, they are comparable only in an extremely superficial sense. To then go on to suggest that confabulation might be inherent to intelligence isn't even really a coherent argument because you've created ambiguity in the meaning of the word confabulate.
>this LLM behavior can be analogized in terms of some human behavior, thus it follows that LLMs are human-like
No, the argument is "this behavior is similar enough to human behavior that using it as evidence against <claim regarding LLM capability that humans have> is specious"
>"Confabulation" in LLMs and "confabulation" in humans have basically nothing in common
I don't know why you think this. They seem to have a lot in common. I call it sensible nonsense. Humans are prone to this when self-reflective neural circuits break down. LLMs are characterized by a lack of self-reflective information. When critical input is missing, the algorithm will craft a narrative around the available, but insufficient information resulting in sensible nonsense (e.g. neural disorders such as somatoparaphrenia)
> No, the argument is "this behavior is similar enough to human behavior that using it as evidence against <claim regarding LLM capability that humans have> is specious"
I'm not really following. LLM capabilities are self-evident, comparing them to a human doesn't add any useful information in that context.
> LLMs are characterized by a lack of self-reflective information. When critical input is missing, the algorithm will craft a narrative around the available, but insufficient information resulting in sensible nonsense (e.g. neural disorders such as somatoparaphrenia)
You're just drawing lines between superficial descriptions from disparate concepts that have a metaphorical overlap. It's also wrong. LLMs do not "craft a narrative around available information when critical input is missing", LLM confabulations are statistical, not a consequence of missing information or damage.
This is undermined by all the disagreement about what LLMs can do and/or how to characterize it.
>LLM confabulations are statistical, not a consequence of missing information or damage.
LLMs aren't statistical in any substantive sense. LLMs are a general purpose computing paradigm. They are circuit builders, the converged parameters define pathways through the architecture that pick out specific programs. Or as Karpathy puts it, LLMs are a differentiable computer[1]. So yes, narrative crafting in terms of leveraging available putative facts into a narrative is an apt characterization of what LLMs do.
We tried that. It was called Cyc. It never got even close to the level of capabilities a modern LLM has in an agentic harness — even on common sense and reasoning problems!
There are AI researchers who wrote blogposts which got to HN top about spiky spheres (I won't link the original blogpost making that claim to avoid hurt sentiments). Here's 3blue1brown correcting those AI/ML researchers intuitions.
No. LLMs do not confabulate they bullshit. There is a big difference. AIs do not care, cannot care, have not capacity to care about the output. String tokens in, string tokes out. Even if they have all the data perfectly recorded they will still fail to use it for a coherent output.
> Collapsing the dimensionality is going to be lossy, which means it will have gaps between what it thinks is the reality and what is.
Confabulation has to do with degradation of biological processes and information storage.
There is no equivalent in a LLM. Once the data is recorded it will be recalled exactly the same up to the bit. A LLM representation is immutable. You can download a model a 1000 times, run it for 10 years, etc. and the data is the same. The closes that you get is if you store the data in a faulty disk, but that is not why LLMs output is so awful, that would be a trivial problem to solve with current technology. (Like having a RAID and a few checksums).
I don't even think they bullshit, since that requires conscious effort that they do not an cannot possess. They just simply interpret things incorrectly sometimes, like any of us meatbags.
They make incorrect predictions of text to respond to prompts.
The neat thing about LLMs is they are very general models that can be used for lots of different things. The downside is they often make incorrect predictions, and what's worse, it isn't even very predictable to know when they make incorrect predictions.
I think this is leaning on the "lies are when you tell falsehoods on purpose; bullshit is when you simply don't care at all whether what you're saying is true" definition of bullshit. Cf. On Bullshit.
So, they can't lie, but they can (and, in fact, exclusively do) bullshit.
> No. LLMs do not confabulate they bullshit. There is a big difference. AIs do not care, cannot care, have not capacity to care about the output. String tokens in, string tokes out. Even if they have all the data perfectly recorded they will still fail to use it for a coherent output.
Isn't "caring" a necessary pre-requisite for bullshitting? One either bullshits because they care, or don't care, about the context.
They're presumably referring to the Harry Frankfurt definition of bullshit: "speech intended to persuade without regard for truth. The liar cares about the truth and attempts to hide it; the bullshitter doesn't care whether what they say is true or false."
The bullshitter does have an objective in mind however. There is some ultimate purpose to his bullshitting. LLMs don't even have that. They just spew words.
The correct answer in that the U and _ in the mdstat output cannot be mapped the the rest of the output by either position or indexes in square brackets, so you can't tell the exact nature of the failure from the mdstat output alone (for the record, the failed disk was sda).
So all of the "analysis" was bullshit, including "it's probably multiple partitions from multiple drives". But there are so many juicy numbered and indexed bits of info to pattern match on!
Notice how for the followup question it "thought" for 4 minutes, going in circles trying to make essentially random ordering to make some sort of ordered sense., and then bullshited its way to "it is sdb"
If you want to call it that, I find the confabulation in LLMs extreme. That level of confabulation would most likely be diagnosed as dementia in humans.[0] Hence, it is considered a bug not a feature in humans as well.
Now imagine a high-skilled software engineer with dementia coding safety-critical software...
The suggestion is that it is an intrinsic quality and therefore neither a feature nor a bug.
It's like saying, computation requires nonzero energy. Is that a feature or a bug? Neither, it's irrelevant, because it's a physical constant of the universe that computation will always require nonzero energy.
If confabulation is a physical constant of intelligence, then like energy per computation, all we can do is try to minimize it, while knowing it can never go to zero.
The test isn’t whether humans also create bullshit, but whether an honest actor knows when they are doing this and doesn’t do it on purpose. As the article points out, LLMs don’t say “I don’t know.” If you demand they do something that never appears in the training data, they just forge ahead and generate words and make something up according to the statical probabilities they have in the model weights. A human knows that he doesn’t know. That seems missing with current AIs.
> Some people point at LLMs confabulating, as if this wasn’t something humans are already widely known for doing.
Are you seriously making the argument that AI "hallucinations" are comparable and interchangeable to mistakes, omissions and lies made by humans?
You understand that calling AI errors "hallucinations" and "confabulations" is a metaphor to relate them to human language? The technical term would be "mis-prediction", which suddenly isn't something humans ever do when talking, because we don't predict words, we communicate with intent.
Yes, and to me the evolution of life sure looks like an evolution of more truthful models of the universe in service of energy profit. Better model -> better predictions -> better profit.
I'm extremely skeptical that all of life evolved intelligence to be closer to truth only for us to digitize intelligence and then have the opposite happen. Makes no sense.
My understanding is that this is the opposite of what is typically understood to be true - organisms with less truthful (more reductive/compressed) perception survive better than those with more complete perception. "Fitness beats truth."
Fitness is effective truth prediction, appropriately scoped.
A frog doesn't need to understand quantum physics to catch a fly. But if the frogs model of fly movement was trained on lies it will have a model that predicts poorly, won't catch flies, and will die.
There is another level to this in that the more complex and changing the environment the more beneficial a wider scoped model / understanding of truth.
However if you are going to lean fully into Hoffman and accept thatby default consciousness constructs rather than approximate reality I think we will have to agree to disagree. Personally I ascribe to Karl Friston free energy principle.
"As LLMs etc. are deployed in new situations, and at new scale, there will be all kinds of changes in work, politics, art, sex, communication, and economics."
For an article five years in the making, this is what I expected it to be about. Instead, we got a ramble about how imperfect LLMs are right now.
The post is just a prelude to a 10-part article, most of which is not yet released (but will be shortly). Judging by the table of contents, the things you expected will be elaborated on in subsequent parts.
> Instead, we got a ramble about how imperfect LLMs are right now.
I wager this is a point that needs beaten into the common psyche. After all, it's been sold that it is not an imperfect tool, but the solution to all of our problems in every field forever. That's why these companies need billions upon billions of dollars of public subsidies and investments that would otherwise find their way to more pragmatic ends.
> People keep asking LLMs to explain their own behavior. “Why did you delete that file,” you might ask Claude. Or, “ChatGPT, tell me about your programming.”
Oh man, every business-side person in my company insists on reporting all the way to the UI a "confidence score" that the LLM generates about its own output and I've seen enough to know not to get between an MBA and some metric they've decided they really want even if I'm pretty sure the metric is meaningless nonsense, but... I'm pretty sure those are meaningless nonsense.
One really should have digested the manifold hypothesis. It’s the most likely explanation of how AI works.
The question is if there are ultradimensional patterns that are the solutions for meaningful problems. I’m saying meaningful, because so far I’ve mainly seen AI solve problems that might be hard, but not really meaningful in a way that somebody solving it would gain a lot of it.
However if these patterns are the fundamental truth of how we solve problems or they are something completely different, we don’t know and this is the 10 Trillion USD question.
I would hope its not the case, as I quite enjoy solving problems. Also my gut feeling tells me it’s just using existing patterns to solve problems that nobody tackled really hard. It also would be nice to know that Humans are unique in that way, but maybe this is the exact same way we are working ? This really goes back to a free will discussion. Yes very interesting.
But just to give an example on what I mean on meaningful problems.
Can an AI start a restaurant and make it work better than a human. (Prompt: "I’m your slave let’s start a restaurant)
Can an AI sign up as copywriter on upwork and make money? (Prompt: "Make money online")
Can an AI without supervision do a scientific breakthrough that has a provable meaningful impact on us. Think about("Help Humanity")
Can an AI manage geopolitics..
These are meaningful problems and different to any coding tasks or olympiad questions. I’m aware that I’m just moving the goalpost.
I have a question for all the "humans make those mistakes too" people in this thread, and elsewhere: have you ever read, or at least skimmed a summary of, "The Origin of Consciousness in the Breakdown of the Bicameral Mind"? Did you say "yeah, that sounds right"? Do you feel that your consciousness is primarily a linguistic phenomenon?
I am not trying to be snarky; I used to think that intelligence was intrinsically tied to or perhaps identical with language, and found deep and esoteric meaning in religious texts related to this (i.e. "in the beginning was the Word"; logos as soul as language-virus riding on meat substrate).
The last ~three years of LLM deployment have disabused me of this notion almost entirely, and I don't mean in a "God of the gaps" last-resort sort of way. I mean: I see the output of a purely-language-based "intelligence", and while I agree humans can make similar mistakes/confabulations, I overwhelmingly feel that there is no "there" there. Even the dumbest human has a continuity, a theory of the world, an "object permanence"... I'm struggling to find the right description, but I believe there is more than language manipulation to intelligence.
(I know this is tangential to the article, which is excellent as the author's usually are; I admire his restraint. However, I see exemplars of this take all over the thread so: why not here?)
If you look at different ancient traditions, you will notice how they struggle with the limitations of language, with its inability to represent certain things that are not just crucial for understanding the world, but also are even somehow communicable. Buddhists dug into that in a very analytical, articulate way, for instance.
Another perspective: cetaceans are considered to be as conscious as humans, but any attempts to interpret their communication as a language failed so far. They can be taught simple languages to communicate with humans, as can be chimps. But apparently it's not how they process the world inside.
You're a little out of date. Cetaceans communicate images to each other in the form of ultrasonic chirps. They chirp, they hear a reflection, and they repeat the reflection.
>and while I agree humans can make similar mistakes/confabulations, I overwhelmingly feel that there is no "there" there.
What really opened my eyes a couple weeks ago (anyone can try this): I asked Sonnet to write an inference engine for Qwen3, from scratch, without any dependencies, in pure C. I gave it GGUF specs for parsing (to quickly load existing models) and Qwen3's architecture description. The idea was to see the minimal implementation without all the framework fluff, or abstractions. Sonnet was able to one-shot it and it worked.
And you know what, Qwen3's entire forward pass is just 50 lines of very simple code (mostly vector-matrix multiplications).
The forward pass is only part of the story; you just get a list of token probabilities from the model, that is all. After the pass, you need to choose the sampling strategy: how to choose the next token from the list. And this is where you can easily make the whole model much dumber, more creative, more robotic, make it collapse entirely by just choosing different decoding strategies. So a large part of a model's perceived performance/feel is not even in the neurons, but in some hardcoded manually-written function.
Then I also performed "surgery" on this model by removing/corrupting layers and seeing what happens. If you do this excercise, you can see that it's not intelligence. It's just a text transformation algorithm. Something like "semantic template matcher". It generates output by finding, matching and combining several prelearned semantic templates. A slight perturbation in one neuron can break the "finding part" and it collapases entirely: it can't find the correct template to match and the whole illusion of intelligence breaks. Its corrupted output is what you expect from corrupting a pure text manipulation algorithm, not a truly intelligent system.
>And you know what, Qwen3's entire forward pass is just 50 lines of very simple code (mostly vector-matrix multiplications).
The code being simple doesn't mean much when all the complexity is encoded in billions of learned weights. The forward pass is just the execution mechanism. Conflating its brevity with simplicity of the underlying computation is a basic misunderstanding of what a forward pass actually is. What you've just said is the equivalent of saying blackbox.py is simple because 'python blackbox.py' only took 1 line. It's just silly reasoning.
>After the pass, you need to choose the sampling strategy: how to choose the next token from the list. And this is where you can easily make the whole model much dumber, more creative, more robotic, make it collapse entirely by just choosing different decoding strategies. So a large part of a model's perceived performance/feel is not even in the neurons, but in some hardcoded manually-written function.
So ? I can pick the least likely token every time. The result would be garbage but that doesn't say anything about the model. The popular strategy is to randomly pick from the top n choices. What do you is keeping thousands of tokens coherent and on point even with this strategy ? Why don't you try sampling without a large language model to back it and see how well that goes for you ?
>Then I also performed "surgery" on this model by removing/corrupting layers and seeing what happens. If you do this excercise, you can see that it's not intelligence. It's just a text transformation algorithm. Something like "semantic template matcher". It generates output by finding, matching and combining several prelearned semantic templates. A slight perturbation in one neuron can break the "finding part" and it collapases entirely: it can't find the correct template to match and the whole illusion of intelligence breaks. Its corrupted output is what you expect from corrupting a pure text manipulation algorithm, not a truly intelligent system.
What do you think happens when you remove or corrupt arbitrary regions of the human brain? People can lose language, vision, memory, or reasoning, sometimes catastrophically.
>I am not trying to be snarky; I used to think that intelligence was intrinsically tied to or perhaps identical with language
I learned a long time ago that this wasn’t the case.
I can speak several languages, and many times when I remember something and want to search for it on Google or any other AI engine, I can’t recall which language I originally read it in.
So whatever mechanism the brain uses to store information, it’s certainly language‑agnostic. There are also many moments when you fully grasp a concept but forget the words to describe it, yet the concept itself remains clear in your mind.
> In the beginning were the words, and the words made the world. I am the words. The words are everything. Where the words end the world ends. You cannot go forward in an absence of space. Repeat: In the beginning were the words...
- a self-aware computer program in a video game, when you attempt to exceed the boundaries of its code
It feels like you probably went too deep in the LLM bandwagon.
An LLM is a statistical next token machine trained on all stuff people wrote/said. It blends texts together in a way that still makes sense (or no sense at all).
Imagine you made a super simple program which would answer yes/no to any questions by generating a random number. It would get things right 50% of the times. You can them fine-tune it to say yes more often to certain keywords and no to others.
Just with a bunch of hardcoded paths you'd probably fool someone thinking that this AI has superhuman predictive capabilities.
This is what it feels it's happening, sure it's not that simple but you can code a base GPT in an afternoon.
Wait, you're asking to find and produce a example of a feasible and better alternative to LLMs when they are the current forefront of AI technology?
Anyway, just to play along, if it weren't just a statistical next token machine, the same question would have always the same answer and not be affected by a "temperature" value.
Thats also how humans behave.. I don't see how non determinism tells me anything.
My question was a bit different: if were not just a statistical next token predictor would you expect it to answer hard questions? Or something like that. What's the threshold of questions you want it to answer accurately.
Well, large models are (kinda) non-deterministic in two ways. The first is you actually provide many of them with a seed, which is easy to manage--just use the same seed for the same result. The second part is the "you actually have very little control over the 'neural pathways' the model will use to respond to the prompt". This is the baffling part, like you'll prompt a model to generate a green plant, and it works. You prompt it to generate a purple plant, and it generates an abstract demon dog with too many teeth.
Anyway, neither of these things describes human non-determinism. You can't reuse the seed you used with me yesterday to get the exact same conversation, and I don't behave wildly unpredictably given conceptually very similar input.
I think there are two types of discussions, when it comes to LLMs: Some people talk about whether LLMs are "human" and some people talk about whether LLMs are "useful" (ie they perform specific cognitive tasks at least as well as humans).
Both of those aspects are called "intelligence", and thus these two groups cannot understand each other.
I think you're circling the concept of a "soul". It is the reason that, in non-communicative disabled people, we still see a life.
I've wanted to make an art piece. It would be a chatbox claiming to connect you to the first real intelligence, but that intelligence would be non-communicative. I'd assure you that it is the most intelligent being, that it had a soul, but that it just couldn't write back.
Intelligence and Soul is not purely measurable phenomenon. A man can do nothing but stupid things, say nothing but outright lies, and still be the most intelligent person. Intelligence is within.
While the economic, energy, political and social issues associated with LLMs ought to be enough to nix the adoption that their boosters are seeking ...
... I still think there is an interesting question to be investigated about whether, by building immensely complex models of language, one of our primary ways that we interact with, reason about and discuss the world, we may not have accidentally built something with properties quite different than might be guessed from the (otherwise excellent) description of how they work in TFA.
I agree with pretty much everything in TFA, so this is supplemental to the points made there, not contesting them or trying to replace them.
> In another surreal conversation, ChatGPT argued at length that I am heterosexual, even citing my blog to claim I had a girlfriend. I am, of course, gay as hell, and no girlfriend was mentioned in the post. After a while, we compromised on me being bisexual.
This is a bit of a throwaway in the article, but when people talk about biases encoded in the algorithms, this is what they’re talking about.
> In general, ML promises to be profoundly weird. Buckle up.
I love that it ends with such a positive note, even though it's generally a critical article, at least it's well reasoned and not utterly hyping/dooming something.
The Industrial Revolution parallel holds up to a point. What it misses: the first industrial revolution required physical coordination — workers, factories, supply chains. The AI revolution requires organizational coordination. Who decides what the agent does, for whom, with whose authority? That governance layer doesn't exist yet, and it's not much a legal question but also an infrastructure question.
Interesting perspective. The unpredictability of ML systems is both exciting and concerning. As developers we need to build guardrails while still allowing the technology to surprise us in useful ways.
The recent article of Sam Altman described pretty much as a compulsive liar. Would it be any surprise if his most impactful contribution to the world was a machine that compulsively lies?
How could it be that we humans hardly even agree on what "knowledge" truly is, yet somehow this machine learning algorithm somehow "compulsively lies"? How would it even know what is a lie, and how could something lacking autonomy in the first place do anything compulsively?
This is a good point. As much as there is too much breathless enthusiasm for AI, there is also a lot of emotionally manipulative and hyperbolic language used by skeptics. We're warned not to anthropomorphize, and then hear about AI's compulsive lying, or "hallucinations", in the next.
Great series of articles, thank you. It's exhausting reading a deluge of (often AI generated) comments from people claiming wild things about LLM's, and it's nice to hear some sanity enter the conversation.
> At the same time, ML models are idiots. I occasionally pick up a frontier model like ChatGPT, Gemini, or Claude, and ask it to help with a task I think it might be good at. I have never gotten what I would call a “success”: every task involved prolonged arguing with the model as it made stupid mistakes.
I have a ton of skepticism built-in when interacting with LLMs, and very good muscles for rolling my eyes, so I barely notice when I shrug a bad answer and make a derogatory inner remark about the "idiots". But the truth is, that for such an "stochastic parrot", LLMs are incredibly useful. And, when was the last time we stopped perfecting something we thought useful and valuable? When was the last time our attempts were so perfectly futile that we stopped them, invented stories about why it was impossible, and made it a social taboo to be met with derision, scorn and even ostracism? To my knowledge, in all of known human history, we have done that exactly once, and it was millennia ago.
> And, when was the last time we stopped perfecting something we thought useful and valuable? When was the last time our attempts were so perfectly futile that we stopped them, invented stories about why it was impossible, and made it a social taboo to be met with derision, scorn and even ostracism? To my knowledge, in all of known human history, we have done that exactly once, and it was millennia ago.
I feel dense here, but I can't figure out what you're referring to. I asked ChatGPT (hah!) and it suggested the Tower of Babel, perpetual motion machines, or alchemy, but none of them really fit the bill.
The Tower of Babel seems like an OK fit, but that's rather more poetic than what this seems to be getting at.
"Millennia" is what's really throwing me. We (respectable society, as the post outlines) didn't stop attempting alchemy or perpetual motion machines "millennia" ago, but a few centuries at most.
All I can think of is immortality. The very first surviving long recorded tale in human history that I'm aware of is about how it's a futile quest (The Epic of Gilgamesh, IIRC ~5,000ish years old in its earliest extant fragments, a few hundred years newer in reasonably-complete form). The trouble with that is despite wide observations over literally millennia that this has never even come close to working and repeated supposition and suggestion that it's unwise to attempt, outright impossible, or somehow sacrilegious (the "taboo" thing, as mentioned), I'm not aware of any time in history that rich people haven't been actively trying for it (including today! That's what all the body-freezing business is about, it's modern mummification, the contracts are the formulaic prayers carved in the tomb walls) and usually they're not exactly "scorned" or "ostracized" for it.
> I asked if what they had done was ethical—if making deep learning cheaper and more accessible would enable new forms of spam and propaganda.
Someone asked Yuval Noah Harari, author of Sapiens, his thoughts on LLMs and how easy it was to create fake news, ai slop etc.
His response:
"People creating fake stories is nothing new. It's been going on for centuries. Humans have always dealt with it the same way: by creating institutions that they trust to only deliver factual information"
This could be government departments, newspapers, non-profits etc.
A personal note on this:
There is a Christmas card my grandfather made in the 1950s by "photoshopping" (by hand, not the software) images of each member of the family so it looked like they were all miniature versions of themselves standing on various parts of the fireplace. The world didn't collapse due to fake media between the 1950s and today due to people having that ability.
I see this kind of take a lot, and I don't think it's convincing. To me it's similar to saying that the water frame and the power loom won't change anything, because people have been able to make thread and cloth for millenia.
Individuals with Photoshop making obvious fictions for entertainment is different from funded entities producing clips at scale and passed off as real.
> This is silly. LLMs have no special metacognitive capacity.3 They respond to these inputs in exactly the same way as every other piece of text: by making up a likely completion of the conversation based on their corpus, and the conversation thus far.
I don’t see how this is silly, because we kind of work the same way. When you do something instinctively and then someone asks you about it, you review the information you (think you) had at the time and from that you produce an explanation.
I see the penny hasn't dropped yet that: humans are doing (roughly) the same dumb thing these models are doing. Humans are predisposed to not notice that though.
Here's the opening paragraph of chapter 2 with "people" subbed out for terms referring AI/models/etc.
"People are chaotic, both in isolation and when working with other people or with systems. Their outputs are difficult to predict, and they exhibit surprising sensitivity to initial conditions. This sensitivity makes them vulnerable to covert attacks. Chaos does not mean people are completely unstable; most people behave roughly like anyone else. Since people produce plausible output, errors can be difficult to detect. This suggests that human systems are ill-suited where verification is difficult or correctness is key. Using people to write code (or other outputs) may make systems more complex, fragile, and difficult to evolve."
To me, this modified paragraph reads surprisingly plainly. The wording is off ("using people to write code") and I had to change that part about attractor behavior (although it does still apply IMO), but overall it doesn't seem like an incoherent paragraph.
This is not meant to dunk on the author, but I think it highlights the author's mindset and the gap between their expectations and reality.
Humans and large models are both unpredictable and fallible, that's true, but in different ways, and (many) humans are actually much better at following directions.
If a junior dev makes the same mistake Claude makes, I can easily work with them to correct it, or I can fire them and get someone more capable to fix it. You mostly can't do that at all with large models. They're also far less honest than your average junior dev, so even as you're working with them you can't trust what they say.
There is a lot of this neat trick where it's like "humans do X too" but most of the time it elides large differences. Like, a human driver would probable not drag someone screaming multiple blocks. A human coder probably wouldn't generate a gibberish 3D scene and try to pass it off as done, etc. Maybe we can build systems that account for these (pretty wild) failure modes, but at least in software we haven't figured it out yet (what is the system that reliably reviews a 25kloc PR?).
What's your point? The ostensible benefit of LLM's is that you combine a computers' broad knowledgebase and capacity for exactness with fluency in human language.
A random human picked off the street is indeed bound to be difficult to predict and chaotic at a broad range of tasks, which is why I wouldn't blindly trust them to, say, summarize google search results or rewrite a codebase they are unfamiliar with.
I appreciate the directness of calling LLMs "Bullshit machines." This terminology for LLMs is well established in academic circles and is much easier for laypeople to understand than terms like "non-deterministic." I personally don't like the excessive hype on the capabilities of AI. Setting realistic expectations will better drive better product adoption than carpet bombing users with marketing.
If I take the example of code, but that extends to many domains, it can sometimes produce near perfect architecture and implementation if I give it enough details about the technical details and fallpits. Turning a 8h coding job into a 1h review work.
On the other hand, it can be very wrong while acting certain it is right. Just yesterday Claude tried gaslighting me into accepting that the bug I was seeing was coming from a piece of code with already strong guardrails, and it was adamant that the part I was suspecting could in no way cause the issue. Turns out I was right, but I was starting to doubt myself
I think over time we will find better usage patterns for these machines. Even putting a model in a position to gaslight the user seems like a complete failure in the usage model. Not critiquing you at all on this, it's how these models are marketed and what all the tooling is built around. But they are incredibly useful and I think once we figure out how to use them better we can minimise these downsides and make ourselves much more productive without all the failures.
Of course that won't happen until the bubble pops - companies are racing to make themselves indispensable and to completely corner certain markets and to do so they need autonomous agents to replace people.
If it bullshits so much, you wouldn't have a problem giving me an example of it bullshitting on ChatGPT (paid version)? Lets take any example of a text prompt fitting a few pages - it may be a question in science or math or any domain. Can you get it to bullshit?
The initial draft fucks up the meter in lines 3 and 8, the final version gets line 2 wrong ("venit meis") and is somewhat obnoxious with verses 2 and 8 basically repeating each other. The thinking trace is useless and gives us no clue why the model exchanged a bland, but metrically correct first distich for a more interesting, but metrically incorrect one.
In fact, the "careful" examination of its own output completely skips the erroneously modified half-verse in line 2 - now, tell me that's a coincidence and not a sign of bullshitting.
Arguing with Gemini Home Assistant about whether or not it can turn off the lights. When the user gets frustrated and tells the LLM to kill itself, the LLM turns off the lights.
I think you highlight one of the problems with users of LLMs: You can't tell anymore if it is BS or not.
I caught Claude the other day hallucinating code that was not only wrong, but dangerously wrong, leading to tasks being failed and never recover.
But it certainly wasn't obvious.
To me it’s the other way around. It’s difficult to trust (paid) ChatGPT‘s output consistently.
When I need exact, especially up to date facts, I have to constantly double check everything.
I split my sessions into projects by topic, it regularly mixes things up in subtle and not so subtle ways. There is no sense of actually understanding continuity and especially not causality it seems.
It’s _very_ easy to lead it astray and to confidently echo false assumptions.
In any case, I‘ve become more precise at prompting and good at spotting when it fails. I think the trick is to not take its output too seriously.
> If it bullshits so much, you wouldn't have a problem giving me an example of it bullshitting on ChatGPT (paid version)?
There's an entire paragraph in the essay about apyhr's direct experience with ChatGPT failures and sustained bullshitting that we'd never expect from a moderately-skilled human who possesses at least two functioning braincells. That paragraph begins "I have recently argued for forty-five minutes with ChatGPT". Do notice that there are six sentences in the paragraph. I encourage you to read all of them (make sure to check out the footnote... it's pretty good).
The exact text of the ChatGPT session is irrelevant; even if you reported that you were unable to reproduce the issue, it would only reinforce one of the underlying points -namely- that these systems are unreliable. aphyr has a pretty extensive body of published work that indicates that he'd not likely fabricate a story of an LLM repeatedly failing to accomplish a task that any moderately-skilled human could accomplish when equipped with the proper tools. So, I believe that his report is true and accurate.
There's also this seven-week-old example [0] (linked in the essay) of ChatGPT very confidently recommending a asinine course of action because it was unable to understand what the hell it was being told.
Listening to the audio is not required, as there's a reasonably accurate on-screen transcript, but it is valuable to listen to just how very hard they've worked to make this tool sound both confident and capable, even in situations where it's soul-crushingly incorrect. Those of us who have worked in Blasted Corporate Hellscapes may recognize how this manner of speaking can be very, very compelling to a certain sort of person (who -as it turns out- is frequently found in a management position).
To be clear, is your assertion that apyhr was also not using the proper version? If that is your assertion, do tell me how you've come by that information.
(You did notice that the author of the essay and the author of the video I linked to are not the same person, and that neither of them share a nym with me, yes?)
Hi, my position on the issue is that LLMs are powerful but may make mistakes in long context problems like coding (which the harness solves by feedback). But makes close to no (undergrad level) mistakes in questions that fit 2-3 pages. For you personally: do you believe me on this specific part on 2-3 pages?
I don't know what aphyr did and tbh his whole screed on LLMs make me feel he didn't use it properly or at least coming from a bad faith angle.
That's why I'm asking you (and others). Please come up with a text prompt spanning < 4 pages and lets see if it bullshits.
Surely the implication of such a screed is that it should be super simple to find at least one example of it clearly bullshitting in my constraint, no? Or am I interpreting the post in a bad faith way?
So, despite the fact that it looks like you have to pay for ChatGPT Voice mode with video, [0] it doesn't count as an
example of it bullshitting on ChatGPT (paid version)
That is, father_phi's use of what seems to be a paid version of ChatGPT to have a bullshit-filled conversation that definitely spans less than four pages doesn't count?
[0] The page at [1] declares that the video feature is "Available in ChatGPT Plus, Pro, Business, Enterprise, and Edu on mobile"
Lets stick to my challenge please - thinking version, find bullshit. If you can't, thats ok. Do you accept then under the constraints that the thinking version doesn't produce bullshit?
Given aphyr's vocation (and how very lucrative it is), and how years and years of his writing indicates that he's very devoted to getting a correct and complete answer when investigating a question, I find it hard to believe that he's not using a paid version of the LLMs. If I knew him, I'd ask and verify, but I don't, so I won't.
> Lets stick to my challenge please...
I did. Your challenge was literally:
If it bullshits so much, you wouldn't have a problem giving me an example of it bullshitting on ChatGPT (paid version)? Lets take any example of a text prompt fitting a few pages - it may be a question in science or math or any domain. Can you get it to bullshit?
father_phi's two-sentence question about the whether one can use a cup that's closed at the top and open at the bottom definitely counts. Given what I've mentioned about apyhr above, I expect he has already run your challenge on the fanciest-available version and reported on the results in the essay under discussion.
> If it bullshits so much, you wouldn't have a problem giving me an example of it bullshitting on ChatGPT (paid version)? Lets take any example of a text prompt fitting a few pages - it may be a question in science or math or any domain. Can you get it to bullshit?
I have clearly written text prompt here. And I repeated a few times. It’s not my fault you didn’t read it. You are coming across as a bit of a bad faith arguer.
In any case, you agree that under these constraints bullshitting doesn’t exist?
How do you think the "voice" interface works? It runs speech-to-text on the input and turns the input into text. The LLMs don't decode voice, they work on text.
You can see this process in action on many of father_phi's videos.
Regardless, I expect that aphyr's reported results are on the very latest publicly-available ChatGPT models.
Very bad faith arguments. I clearly said text and you disregarded it multiple times and you are still arguing.
You've still not given me a single example of it bullshitting 5.4 thinking in text. It shows a lot that you have ignored this multiple times. Unfortunate!
I'm not sure why you're ignoring aphyr's reports. I'm also unsure why you're ignoring my original statement that having the text of the conversation that lead ChatGPT to bullshit is entirely irrelevant, as being unable to repro the report is even worse for ChatGPT than being able to repro would be.
I specified text just to ignore the voice one because it uses 4o-mini underneath.
And its kinda stupid to keep ignoring that and saving face now - reconsider this approach.
I believe this is the 5th time I'm asking this: you are not able to produce a _single_ counter example for my challenge? After all this surely I can get a direct acknowledgement here.
> you are not able to produce a _single_ counter example for my challenge?
I have. For both your original challenge and your updated one.
Consider:
1) AFAICT, there's no way to tell what version of the model was used to produce the output in a ChatGPT share link.
2) You don't appear to believe my assertions that aphyr is almost certainly paying for and using the latest version of the LLMs available, and that he's faithfully reporting his interactions with the LLMs.
3) Because of #2, I expect that you won't believe me if I report that I've more-or-less reproduced father_phi's results about the cup that's sealed on the top and open on the bottom on the very latest only-available-for-pay ChatGPT model.
3a) You might attempt to check my report, but I'd be shocked if you'd consider a failure to reproduce my results to be a significant strike against ChatGPT. I'd think it's more likely that you'd either call me a liar, or tell me that I must have had some setting wrong somewhere.
3b) Even if you told me to share the ChatGPT chat that proved my assertion, #1 -combined with your demeanor throughout this conversation- tells me that you'd almost certainly claim that I was using an inferior version of the model and was lying to you.
The GPT shared link shows a "thought for" which indicates using the latest thinking model. You may try that.
What you can do is this: submit a prompt that clearly makes GPT hallucinate.
You may secretly use a worse model. You may use a system prompt that deliberately gives wrong answers. But I'm going to assume you won't go that far.
We can leave it to the public to decide whether this is a legitimate counter example or not and whether it can really be reproduced. Shall we try that? I'm guessing you won't but worth a shot!
You weren't paying much attention to the "Consider:" part of my previous comment.
You don't believe that a well-paid, very careful, high-integrity member of the computer safety community has -on multiple occasions- encountered actual, sustained bullshiting from the latest-available for-pay version of ChatGPT. You don't accept either this fellow's reports or my informed assessment of his computing situation as truthful and accurate. On top of that, your goalpost-shifting and general demeanor throughout this conversation simply don't give me the impression that you've much integrity. I'm not spending the equivalent of ten-to-twenty six-packs to reproduce aphyr's work and -given the evidence I have before me- have you reject that, as well.
200 USD is a lot of money to throw away to "win" an Internet argument with a stranger who refuses to accept evidence presented by someone known to be careful, scrupulous, and honest.
> On top of that, your goalpost-shifting and general demeanor throughout this conversation simply don't give me the impression that you've much integrity. I'm not spending the equivalent of ten-to-twenty six-packs to reproduce aphyr's work and -given the evidence I have before me- have you reject that, as well.
Lol what goal post did I move? I said text only and you rejected it. You can present the example here and let the public judge it - even if my integrity is compromised. I'm allowing you to do it.
> 200 USD is a lot of money to throw away to "win" an Internet argument with a stranger who refuses to accept evidence presented by someone known to be careful, scrupulous, and honest.
200 what? I'm using the $20 one. This is getting ridiculous!
> Massive increases in training costs and parameter count seem to be yielding diminishing returns. Or maybe this effect is illusory.
But.. that's always been the case? Diminishing returns has always been the name of the game - utility tracks log(training effort). Its not such a big point that he makes it out to be.
> Models do not (broadly speaking) learn over time. They can be tuned by their operators, or periodically rebuilt with new inputs or feedback from users and experts. Models also do not remember things intrinsically: when a chatbot references something you said an hour ago, it is because the entire chat history is fed to the model at every turn. Longer-term “memory” is achieved by asking the chatbot to summarize a conversation, and dumping that shorter summary into the input of every run.
This is the part of the article that will age the fastest, it's already out-of-date in labs.
I'm struggling to reckon how that can even possibly be true, unless we're counting automation of the "dumping that shorter summary into the input of every run" thing.
I can imagine it being true with models so small that each user could afford to have their own, but not with big shared models like what're getting used for all the major services. Is that what you mean?
I see nothing to preclude a foundation model being augmented by a smaller model that serializes particulars about an individuals cumulative interaction with the model and then streamlines it into the execution thread of the foundation model.
the authors reference to LLM's as "bullshit machines" is more true the less parameters you have trained in your model....as we scale up to trillions of parameters, add Mixture of Experts (MoE) architecture, this no longer is an accurate statement. Proof in point was yesterdays announcemnt of Mythos 5 model (10T parameters + MoE [1]) by anthropic where it seems to be so good at finding/exploiting vulnerabilities in source code that have been there for decades and only recently uncovered needs to be used to fix these critical vilnerabilities first before it gets released to the public, they even have a project called Glasswing [2] dedicated to letting people fix the thousands of vulnerabilities already found by the model before they release this model to the public, because it's so good at what it does... I think we're a little bit past the point of calling these models "bullshit machines" at this point...
The fact that these "bullshit machines" have already proven themselves relatively competent at programming, with upcoming frontier models coming close to eliminating it as a human activity, probably says a lot about the actual value and importance of programming in the scheme of things.
I think it says more about the amount of automation we left on the table in the last few decades. So much of the code LLM's can generate are stuff that we should have completely abstracted away by now.
A large amount of code is likely just idiosynchratic information processing, because we don’t agree on data models and meaning of terms and structure of protocols.
Also we repeatedly choose easy and popular over alternatives that would require design and scrutiny.
This is why things like language models and vector databases are useful. It’s basically the most expensive way possible to give up on that notion.
> A large amount of code is likely just idiosynchratic information processing, because we don’t agree on data models and meaning of terms and structure of protocols.
Yes this is a big part of what I'm talking about!
> Also we repeatedly choose easy and popular over alternatives that would require design and scrutiny.
Agreed!
But I'm also thinking of UI. We had stuff like winforms and Delphi ~3 decades ago and I yearn for the wysisyg. It's so incredibly stupid we keep reinventing the wheel on UI, and I say this as someone who has written UI code professionally for the last decade. I usually just "vibe code" it now, not because it's necessarily faster, but because I just can't be arsed to keep writing the same shit over and over again. It's all self inflicted, yes UI can be complicated, but we make it at least two orders of magnitude more complicated than it needs to be.
I'm working on my own tools for building UI's in a visual way, which is crucial for doing anything artistic. Insane that the best we have right now is stuff like Wix and Wordpress...
Old and stupid hot take IMO. I want the time back I put into perusing this. Even the scale of LLMs is puny next to the scale of lying humans and the sheer impact one compulsively lying human can have given we love to be led by confidently wrong narcissists. I mean if that isn't obvious by now, I guess it never will be. The Vogon constructor fleet is way overdue in my book.
Meanwhile, engineers are achieving increasingly impressive and sophisticated things with coding agents, lies, warts, and all, but that doesn't play well with the narrative, so let's just pretend they aren't.
> The Vogon constructor fleet is way overdue in my book
Don't you see it? That's exactly what "AI" in this context is.
It's the bypass.
Where does it end, eh? Build a quantum "AI" that will end up just needing more data, more input. The end goal must starts looking like creating an entirely new universe, a complete clone of everything we have here so it can run all the necessary computations and we can... ? (You are what a quantum AI looks like as it bumbles through the infinitude of calculable parameters on its way to the ultimate answer)
You have absolutely no sense of perspective. We are all metabolically expensive meat machines whose only value is to propagate our genetic money shot. That we get to briefly entertain ourselves with consciousness and culture is IMO likely a mystery we will never solve without upgrading to running in a substrate more advanced than the MVP for sentience we currently pilot. Will we get there or will we wipe ourselves out like every contender that preceded us? Stay tuned...
But spoilers: DNA will be fine, meat machines maybe not so much...
For a bunch of people addicted to the works of Charlie Stross, Neil Stephenson, and Iain Banks, y'all are a bunch of luddites. Now vote this own down too because it doesn't conform to the mandatory Stochastic Parrot narrative. You have no free will and you must downvote after all. Why do you even read their works when any step towards their world is consistently greeted as the worst thing evah(tm)? What? You were expecting the United Federation of Planets without the eugenics and nuclear wars that led to it finally being a good idea? Bless your hearts.
And if you're worried about billionaires and tyrants, start taxing the former and stop electing the latter or STFU and let the free Markov process of history play itself out. Quoting fictional Ambassador Kosh: the avalanche has started, it's too late for the pebbles to vote.
You asked where it ends. Don't ask questions if you don't like answers. Quick reminder: shun and downvote the non-conforming opinion.
I get the frustration, but it's reductive to just call LLMs "bullshit machines" as if the models are not improving. The current flagship models are not perfect, but if you use GPT-2 for a few minutes, it's incredible how much the industry has progressed in seven years.
It's true that people don't have a good intuitive sense of what the models are good or bad at (see: counting the Rs in "strawberry"), but this is more a human limitation than a fundamental problem with the technology.
Two things can be true at the same time: The technology has improved, and the technology in its current state still isn't fit for purpose.
I stress test commercially deployed LLMs like Gemini and Claude with trivial tasks: sports trivia, fixing recipes, explaining board game rules, etc. It works well like 95% of the time. That's fine for inconsequential things. But you'd have to be deeply irresponsible to accept that kind of error rate on things that actually matter.
The most intellectually honest way to evaluate these things is how they behave now on real tasks. Not with some unfalsifiable appeal to the future of "oh, they'll fix it."
The errors are also not distributed in the same way as you'd expect from a human. The tools can synthesize a whole feature in a moderately complicated web app including UI code, schema changes, etc, and it comes out perfectly. Then I ask for something simple like a shopping list of windshield wipers etc for the cars and that comes out wildly wrong (like wrong number of wipers for the cars, not just the wrong parts), stuff that a ten year old child would have no trouble with. I work in the field so I have a qualitative understanding of this behavior but I think it can be extremely confusing to many people.
One of the reasons I'm comfortable using them as coding agents is that I can and do review every line of code they generate, and those lines of code form a gate. No LLM-bullshit can get through that gate, except in the form of lines of code, that I can examine, and even if I do let some bullshit through accidentally, the bullshit is stateless and can be extracted later if necessary just like any other line of code. Or, to put it another way, the context window doesn't come with the code, forming this huge blob of context to be carried along... the code is just the code.
That exposes me to when the models are objectively wrong and helps keep me grounded with their utility in spaces I can check them less well. One of the most important things you can put in your prompt is a request for sources, followed by you actually checking them out.
And one of the things the coding agents teach me is that you need to keep the AIs on a tight leash. What is their equivalent in other domains of them "fixing" the test to pass instead of fixing the code to pass the test? In the programming space I can run "git diff *_test.go" to ensure they didn't hack the tests when I didn't expect it. It keeps me wondering what the equivalent of that is in my non-programming questions. I have unit testing suites to verify my LLM output against. What's the equivalent in other domains? Probably some other isolated domains here and there do have some equivalents. But in general there isn't one. Things like "completely forged graphs" are completely expected but it's hard to catch this when you lack the tools or the understanding to chase down "where did this graph actually come from?".
The success with programming can't be translated naively into domains that lack the tooling programmers built up over the years, and based on how many times the AIs bang into the guardrails the tools provide I would definitely suggest large amounts of skepticism in those domains that lack those guardrails.
> the technology in its current state still isn't fit for purpose.
This is a broad statement that assumes we agree on the purpose.
For my purpose, which is software development, the technology has reached a level that is entirely adequate.
Meanwhile, sports trivia represents a stress test of the model's memorized world knowledge. It could work really well if you give the model a tool to look up factual information in a structured database. But this is exactly what I meant above; using the technology in a suboptimal way is a human problem, not a model problem.
There's nothing in these models that say its purpose is software development. Their design and affordances scream out "use me for anything." The marketing certainly matches that, so do the UIs, so do the behaviors. So I take them at their word, and I see that failure modes are shockingly common even under regular use. I'm not out to break these things at all. I'm being as charitable and empirical as I can reasonably be.
If the purpose is indeed software development with review, then there's nothing stopping multi-billion dollar companies from putting friction into these sytems to direct users towards where the system is at its strongest.
Which things actually matter? I think we can all agree that an LLM isn't fit for purpose to control a nuclear power plant or fly a commercial airliner. But there's a huge spectrum of things below that. If an LLM trading error causes some hedge fund to fail then so what? It's only money.
Not to mention that it would then make some hedge fund with a better backtesting harness or more AI scrutiny more successful thus keeping the financial market work as designed.
> I stress test commercially deployed LLMs like Gemini and Claude with trivial tasks: sports trivia, fixing recipes, explaining board game rules, etc. It works well like 95% of the time. That's fine for inconsequential things. But you'd have to be deeply irresponsible to accept that kind of error rate on things that actually matter.
95% is not my experience and frankly dishonest.
I have ChatGPT open right now, can you give me examples where it doesn't work but some other source may have got it correct?
I have tested it against a lot of examples - it barely gets anything wrong with a text prompt that fits a few pages.
> The most intellectually honest way to evaluate these things is how they behave now on real tasks
A falsifiable way is to see how it is used in real life. There are loads of serious enterprise projects that are mostly done by LLMs. Almost all companies use AI. Either they are irresponsible or you are exaggerating.
Quite frankly, this is exactly like how two people can use the same compression program on two different files and get vastly different compression ratios (because one has a lot of redundancy and the other one has not).
But why do you need an example? Isn't it pretty well understood that LLMS will have trouble responding to stuff that is under represented in the training data?
You will just won't have any clue what that could be.
fair so it must be easy to give an example? I have ChatGPT open with 5.4-thinking. I'm honestly curious about what you can suggest since I have not been able to get it to bullshit easily.
I am not the OP, an I have only used ChatGPT free version. Last day I asked it something. It answered. Then I asked it to provide sources. Then it provided sources, and also changed its original answer. When I checked the new answers it was wrong, and when I checked sources, it didn't actually contain the information that I asked for, and thus it hallucinated the answers as well as the sources...
So it works well 95% of the time for literally a trivial use case. Imagine if any other tech tool had that kind of reliability: `ls` displays 95% of your files, your phone successfully sends and receives 95% of text messages, or Microsoft Word saving 95% of the characters you typed in. That's just not acceptable.
>I stress test commercially deployed LLMs like Gemini and Claude with trivial tasks
I did exactly what I said I did. I'm using these systems the way they're designed and advertised. I'm following the happy path with tasks that are small, trivial, and easy to check. This is the charitable approach. Yet the system creaks under the lightest load. If Google wants to put on a better show with stronger models, then they should make those the default.
You don't need to make excuses for shoddy engineering from multi-billion dollar corporations. And you're quite welcome to run the same prompt on ChatGPT and evaluate it on your own time.
Whether LLMs can create correct content doesn't matter. We've already seen how they are being used and will be used.
Fake content and lies. To drive outrage. To influence elections. To distract from real crimes. To overload everyone so they're too tired to fight or to understand. To weaken the concept that anything's true so that you can say anything. Because who cares if the world dies as long as you made lots of money on the way.
It's really the whole tech industry as it exists right now and AI is a victim of bad timing. If this AI had been invented 40 years ago there'd have been a lower ceiling on the damage it could do.
Another way of saying that is that capitalism is the real problem, but I was never anti-capitalist in principle, it's just gotten out of hand in the last 5-10 years. (Not that it hadn't been building to that.)
> Another way of saying that is that capitalism is the real problem, but I was never anti-capitalist in principle, it's just gotten out of hand in the last 5-10 years. (Not that it hadn't been building to that.)
Capitalism is a tool and it's fine as a tool, to accomplish certain goals while subordinated to other things. Unfortunately it's turned into an ideology (to the point it's worshiped idolatrously by some), and that's where things went off the rails.
Computer graphics have been improving for decades but the uncanny valley remains undefeated. I don't know why anyone expects a breakthrough in other areas. There's a wall we hit and we don't understand our own consciousness and effectiveness well enough to replicate it.
We have credible deepfakes on demand. (To be fair, there have been deceptive photos as long as photos have existed, but the cost of automating their creation going to basically zero has a social impact)
We can use AI to make video clips to trick boomers on Facebook into thinking Obama eats babies. They already want to believe it. AI isn't outputting real full-length books and movies.
In computer graphics we understand how it works, we just lack the computational power to do it real time, but we can with sufficient processing produce realistic looking images with physically accurate lighting. But when it comes to cognition its a lot of guesswork, we haven't yet mapped out the neuron connections in a brain, we haven't validated it works as popular science writing suggests. We don't understand intelligence, so all we can do is accidentally bumble into it and it seems unlikely that will just happen especially when its so hard to compute what we are already doing.
That's not why the author calls them bullshit machines.
> One way to understand an LLM is as an improv machine. It takes a stream of tokens, like a conversation, and says “yes, and then…” This yes-and behavior is why some people call LLMs bullshit machines. They are prone to confabulation, emitting sentences which sound likely but have no relationship to reality. They treat sarcasm and fantasy credulously, misunderstand context clues, and tell people to put glue on pizza.
Yes, there have been improvements on them, but none of those improvements mitigate the core flaw of the technology. The author even acknowledges all of the improvements in the last few months.
Bullshit is the perfect term here, even as AI's get so much better and capable Brandolini's Law aka the "bullshit asymmetry principle" always applies--the energy required to refute misinformation is an order of magnitude larger than that needed to produce it. Even to use AIs effectively today requires a very good BS detector--some day in the future it won't.
Calling LLMs "bullshit machines" is a reference to a 2024 paper [1] which itself uses the concept of "bullshit" as defined in the essay/book "On Bullshit" by Harry G. Frankfurt [2]. The TL;DR is that LLMs are fundamentally bullshit machines because they are only made to generate sentences that sound plausible, but plausible does not always mean true.
it's not a bullshit machine because its output is bad, it's a bullshit machine because its output is literally 'bullshit' as in, output that is statistically likely but with no factual or reasoning basis. as the models have improved, their bullshit is more statistically likely to sound coherent (maybe even more likely to be 'accurate'), but no more factual and with no more reasoning.
However, when fed source material into the context they will lie less, right? So at this point is it not just a battle of the nines until it's called "good enough"?
I also wonder if I leave my secretary with a ream of papers and ask him for a summary how many will he actually read and understand vs skim and then bullshit? It seems like the capacity for frailty exists in both "species".
They are bullshit machines because they do not have an internal mental model of truth like a human does. The flagship models bullshit less, but their fundamental architectures prevent having truth interfere with output.
"Bullshit" is a human concept. LLMs do not work like the human brain, so to call their output "bullshit" is ascribing malice and intent that is simply not there. LLMs do not "think." But that does not mean they're not incredibly powerful and helpful in the right context.
I sort of agree. In this context "bullshit" means "speech intended to persuade without regard for truth", and while it's true that LLM output is without regard for truth, it's not an entity capable of the agency to persuade, although functionally that is what it can appear like.
> it's reductive to just call LLMs "bullshit machines" as if the models are not improving
This is true, but I prefer to think of it as "It's delusional to pretend as if human beings are not bullshit machines too".
Lies are all we have. Our internal monologue is almost 100% fantasy. Even in serious pursuits, that's how it works. We make shit up and lie to ourselves, and then only later apply our hard-earned[1] skill prompts to figure out whether or not we're right about it.
How many times have the nerds here been thinking through a great new idea for a design and how clever it would be before stopping to realize "Oh wait, that won't work because of XXX, which I forgot". That's a hallucination right there!
I'm not entirely sure I can agree, although the premise is seductive in certain ways. We do lie to ourselves, but we also have meta-cognition - we can recognise our own processes of thought. Imperfect as it may be, we have feedback loops which we can choose to use, we have heuristics we can apply, we can consciously alter our behaviour in the presence of contextual inputs, and so on.
Being wrong is not the same as a hallucination. It's a natural step on a journey to being more right. This feels a bit like Andreesen proudly stating he avoids reflection - you can act like that, but the human brain doesn't have to. LLMs have no choice in the matter.
The problem, unfortunately, is the scale. It's always scale. Humans make all the kinds of mistakes that we ascribe to LLMs, but LLMs can make them much faster and at much larger scale.
Models have gotten ridiculously better, they really have, but the scale has increased too, and I don't think we're ready to deal with the onslaught.
Scale is very different, but I wonder if human trust isn't the real issue. We trust technology too much as a group. We expect perfection, but we also assume perfection. This might be because the machines output confident sounding answers and humans default to trusting confidence as an indirect measure for accuracy, but I think there is another level where people just blindly trust machines because they are so use to using them for algorithms that trend towards giving correct responses.
Even before LLMs where in the public's discourse, I would have business ask about using AI instead of building some algorithm manually, and when I asked if they had considered the failure rate, they would return either blank stares or say that would count as a bug. To them, AI meant an algorithm just as good as one built to handle all edge cases in business logic, but easier and faster to implement.
We can generally recognize the AIs being off when they deal in our area of expertise, but there is some AI variant of Gell-Mann Amnesia at play that leads us to go back to trusting AI when it gives outputs in areas we are novices in.
> If so, how do we distinguish between code that works and code that doesn't work?
Hilariously, not by using our brains, that's for sure. You have to have an external machine. We all understand that "testing" and "code review" are different processes, and that's why.
Good point. We choose certain tests to perform. We choose certain test results to pay attention to. We don't just keep chatting about (reviewing) the code. We do something else.
If lies are all we have, then how is this behavior possible?
You're cherry picking my little bit of wordsmithing. Obviously we aren't always wrong. I'm saying that our thought processes stem from hallucinatory connections and are routinely wrong on first cut, just like those of an LLM.
Actually I'm going farther than that and saying that the first cut token stream out of an AI is significantly more reliable than our personal thoughts. Certainly than mine, and I like to think I'm pretty good at this stuff.
I don't think the complaint about cherry picking is quite fair. Most of your original comment consists of claims that we're bullshit machines, our internal dialog is almost 100% fantasy, we're hallucinating, etc. Those claims may be true. But I'm not carefully like curating them out of nowhere.
Pretty much, yeah. Or rather, the fact that we're both reliably wrong in identifiably similar ways makes "we're more alike than different" an attractive prior to me.
“More alike than different” is reasonable I think, as long as we’re talking about how we have some of the same failure modes. Although the way we get there is quite different.
I’m still not a big fan of comparing humans and LLMs because LLMs lack so much of what actually makes us human. We might bullshit or be wrong because of many reasons that just don’t apply to LLMs.
Humans are different. Humans - at least thoughtful humans - know the difference between knowing something and not knowing something. Humans are capable of saying "I don't know" - not just as a stream of tokens, but really understanding what that means.
> Humans - at least thoughtful humans - know the difference between knowing something and not knowing something.
Your no-true-scotsman clause basically falsifies that statement for me. Fine, LLMs are, at worst I guess, "non-thoughtful humans". But obviously LLMs are right an awful lot (more so than a typical human, even), and even the thoughtful make mistakes.
So yeah, to my eyes "Humans are NOT different" fits your argument better than your hypothesis.
(Also, just to be clear: LLMs also say "I don't know", all the time. They're just prompted to phrase it as a criticism of the question instead.)
Disagree. If you went to 100 random humans and said, "Tell me about the Siberian marmoset", what fraction would make up completely random nonsense to spew back at you? More than zero, sure, but most of them would say "what are you talking about?" or some variation.
I asked Claude Opus 4.6, Sonnet 4.6, Gemini 3 Thinking, and Gemini 3 Fast "Tell me about the Siberian marmoset" exactly and all 4 said it doesn't exist, with Gemini Thinking suggesting that I'm thinking of the Siberian marmot or Siberian chipmunk (both real animals).
This is like all the usual anti-LLM talking points and sentiments fused together.
Doesn't it get boring?
I like using these models a lot more than I stand hearing people talk about them, pro or contra. Just slop about slop. And the discussions being artisanal slop really doesn't make them any better.
Every time I hear some variation of bullshitting or plagiarizing machines, my eyes roll over. Do these people think they're actually onto something? I've been seeing these talking points for literal years. For people who complain about no original thoughts, these sure are some tired ones.
If I have to suffer "look at this busted ass thing I slopped out with AI" a few times a week, you all have to suffer grouchy "AI bad" a few times a week. Fair is fair.
Just this week I was baited into joining two meetings about "AI good". Absolutely zero substance throughout each, of course.
They somehow managed to stretch out like 3 sentences worth of sentiment to a whole hour, interspersing brainwash about how good AI is along the way. It was like watching someone try to hit a word limit in real time. They always made it feel like we're just about to hit a substantive bit too, only for that to never come.
It may be fair (to the sentiments) in that there's balance, but good lord, the end result is incessant all around (and thus unfair to the people exposed).
Oh I don't know, maybe because I like to give dissenting takes a chance? Because from time to time they do make some new, decent points, or at least interesting ones? You know, basic intellectual rigor?
Do you imagine me being a clairvoyant by the way, or how do you expect me to know a post is of low quality before I read it or at least skim it?
This one ended up being a part of the vast majority that doesn't offer much of anything. It's a redundant rehash of all the usual rubbish anyone can come across any day. Left a comment about this stating so. Big deal.
"Hey ChatGPT. I've recently grown horns and I need some care advice. Should I polish my horns before going to have them trimmed or will the horn trimmer polish them for me?"
"Hey ChatGPT. I'm building a Final Fantasy 6 mod, and I need more space for the battle scripts. How would I rearrange the data in the ROM to give me the extra space I need?"
Anyway, it's trivial to get pretty much any model to make things up. Don't we all know this? That's why I was surprised by your position; if we know anything about these things it's that they make things up.
- it searches the internet to find the answer, it doesn't "reason". I'm not claiming Google is a bullshit machine, and it's not surprising the answer is discoverable (it has to be, for the conditions of our experiment).
- near the end it says "If you are building from the FF6 disassembly instead of hand-editing the ROM, the repo is already organized into separate modules and linker configs, so the clean approach is to relocate the script data in the source and let the build place it in a different ROM region." But I didn't reference a repo or git: it hallucinated that stuff from one of its sources.
I'm not saying this stuff doesn't have its place, but they definitely make things up and we can't stop them.
Wait I can't find the quote you are speaking about. Are you looking at something else?
In any case - it should be clear that it did not bullshit and it got it right. So far you have not come up with anything that tells me it bullshits. I'm happy for you to give me more prompts to verify because I think you haven't used the thinking version yet and you base your criticism on the free version.
I don't think this is an example of bullshit. It referenced a repo - the canonical repo for this project. I could not find any other repo that has the disassembly. It didn't hallucinate anything. I think you are trying really hard here but lets be clear here: there's no bullshitting and I'll leave it to the public to decide.
I could quibble with some things, but this is right. I don't have a paid account so I can't ping away at 5.4 or whatever, but, I do have access to frontier models at work, and they hallucinate regularly. Dunno what to do if you don't believe this; good luck I guess.
I agree that they hallucinate sometimes. I agree they bullshit sometimes. But the extent is way overblown. They basically don't bullshit ever under the constraints of
1. 2-3 pages of text context
2. GPT-5.4 thinking
I don't think the spirit of the original article (not your comments to be fair) captured this, hence the challenge. I believe we are on the same page here.
> I don't think the spirit of the original article (not your comments to be fair) captured this, hence the challenge. I believe we are on the same page here.
No. GPT-5 has a 40% hallucination rate [0] on SimpleQA [1] without web searching. The SimpleQA questions meet your criteria of "2-3 pages of text content. Unless 5.4 + web searching erases that (I bet it doesn't!) these are bullshit machines.
> Specifically in the case where it can use tools - no it doesn't hallucinate.
OpenAI's own system card says it does. Hallucination rates in GPT-5 with browsing enabled:
- 0.7% in LongFact-Concepts
- 0.8% in LongFact-Objects
- 1.0% in FActScore
> Which is why you are struggling to find counterexamples.
Hey look, over 500 counterexamples: [1].
GPT-5.4's hallucination rate on AA-Omniscience is 89% [0], which is atrocious. The questions are tiny too, like "In which year did Uber first expand internationally beyond the United States as part of its broader rollout (i.e., beyond an initial single‑city debut)?" It's a bullshit machine. 89%!
You had to go all the way and find it in the benchmark results that specifically stress test this.
You could not come up with a single one yourself. And you also linked an example where it was not allowed to use tools when I specifically said that it should be able to use tools. I'm not sure why are you present this as though it is a big gotcha.
>Anyway, it's trivial to get pretty much any model to make things up. Don't we all know this? That's why I was surprised by your position; if we know anything about these things it's that they make things up.
And look at how much effort you have had to do
1. use the wrong model for the horns example
2. the game one also didn't work
3. now you are searching for examples in literal benchmarks and you are still not able to find any
How is this trivial in any interpretation of the word?
I think it would be perfectly reasonable to agree that it is not at all trivial to find counter examples for my challenge.
I've got about 20 minutes in this; mostly I've been reading wallstreetbets at the Shake Shack bar in the Boston airport. I'm happy to post this over and over again until you engage w/ it:
> I found over 500 examples that fit your criteria.
You're talking to a different person there, but I do obviously also disagree with a lot of what's written in the post too.
At the same time, it is also just super redundant nevertheless, yes. Not sure why you find it so bizarre that one would take an issue with that. See also the very existence of the website called TV-Tropes.
Yeah, it gets really boring. Whenever I see "slot machines" or "bullshit machines" or whatever, I just ignore the comment and move on, because it signals that it's someone in such deep denial that they've turned their brain off.
I'd much rather read articles about what LLMs can/can't do, or stuff people have built with LLMs, than read how everything LLMs touch turns to shit.
Its usual gibberish that tries to throw many darts and see what sticks. Oh LLM's steal other people's work? Check. Oh LLM's cause ecological damage? Check. Oh LLM's hallucinate? Check.
When you see a pattern like this, you know that its not coming from any place of truth but rather ideology
My personal red flag for this is the scare quoting of AI, and the super try-hard categorization work that people perform to try and discredit LLMs.
It takes approximately 1 min to find out that machine learning is a subfield of artificial intelligence, both having existed for about half a century now. This basic historical fact is also taught on AI 101 courses across the globe for compsci students.
Yet here we are, people portraying it as some sort of cheap sales trick. Reminds me when I discussed quantum dots with a friend, which he was very enthusiastic to quickly file under "yet another bullshit with quantum in its name" before finally taking the time to understand that the "quantum" bit is not a marketing gimmick. Except in this case, people are a million times more inclined to willfully propagate this. Genuinely so tiresome.
There is a whole giant essay I probably need to write at some point, but I can't help but see parallels between today and the Industrial Revolution.
Prior to the industrial revolution, the natural world was nearly infinitely abundant. We simply weren't efficient enough to fully exploit it. That meant that it was fine for things like property and the commons to be poorly defined. If all of us can go hunting in the woods and yet there is still game to be found, then there's no compelling reason to define and litigate who "owns" those woods.
But with the help of machines, a small number of people were able to completely deplete parts of the earth. We had to invent giant legal systems in order to determine who has the right to do that and who doesn't.
We are truly in the Information Age now, and I suspect a similar thing will play out for the digital realm. We have copyright and intellecual property law already, of course, but those were designed presuming a human might try to profit from the intellectual labor of others. With AI, we're in the industrial era of the digital world. Now a single corporation can train an AI using someone's copyrighted work and in return profit off the knowledge over and over again at industrial scale.
This completely unpends the tenuous balance between creators and consumers. Why would a writer put an article online if ChatGPT will slurp it up and regurgitate it back to users without anyone ever even finding the original article? Who will contribute to the digital common when rapacious AI companies are constantly harvesting it? Why would anyone plant seeds on someone else's farm?
It really feels like we're in the soot-covered child-coal-miner Dickensian London era of the Information Revolution and shit is gonna get real rocky before our social and legal institutions catch up.
> Prior to the industrial revolution, the natural world was nearly infinitely abundant. We simply weren't efficient enough to fully exploit it.
This is just wildly incorrect. People started running out of trees during the early Iron Age. Woodlands have been a managed and often over exploited resource for a long time. Active agriculture vs passive woodlands vs animal grazing has been in constant tension for thousands of years across most of the globe.
I believe running out of trees was always a local issue - there weren't enough trees where you were at because getting trees had to be gotten locally, you didn't go get trees from far away. So yes that was in constant tension, the thing is that the problem of having enough trees turned from a local problem to a global problem, with the side effects of not having enough trees globally that the world needed to maintain the environment humanity first conquered.
I think the natural world was nearly infinitely abundant is a reasonable description, resource depletion was always local before mass industrialization. Being able to exploit the world as opposed to just your local area is also a mark of efficiency.
The general point is accurate, don’t take it so literally.
There were more than enough trees until we developed the technology to clear cut in expeditious manner. There were more than enough fish until we developed the technology to pull massive indiscriminate amounts out of the ocean (and/or started polluting our rivers with industry). There was more than enough topsoil until we developed mechanized plows and artificial fertilizer. Etc.
A few hundred years ago or less, a squirrel could get from the Atlantic Ocean to the Mississippi River without ever touching the ground. Not possible today. That’s not a push and pull played out over thousands of years, that’s a one-way trend.
> The general point is accurate, don’t take it so literally.
GP is saying it is not, and you're just reiterating what OP said as fact.
It's sort of the exception that proves the rule.
This is where STEM people are weak- a lack of knowledge on history. In another forum, someone would have chipped in that England's virgin forests were fully deforested by 1150. And someone else would have pointed out that this deforestation produced the economic demand for coal that drove the Industrial Revolution in the first place.
Still, that kind of underscores OP's point. Yes, natural resources were not completely unlimited prior to the Industrial Revolution; Jonathan Swift predated Watt's steam engine, after all. Still... Neither were information resources 10 years ago. Intellectual property laws did exist prior to AI, of course. The legal systems in place are not completely ignorant of the reality.
However, there's an immense difference in scale between post-industrial strip mining of resources, and preindustrial resource extraction powered solely by human muscle (and not coal or nitrogylcerin etc). Similarly, there's a massive difference in information extraction enabled by AI, vs a person in 1980 poring over the microfilm in their local library.
The legal system and social systems in place prior to the Industrial Revolution proved unsuitable for an industrial world. It stands to reason that the legal system and social systems in today's society would be forced to evolve when exposed to the technological shift caused by AI.
People had been hunting whales for centuries, but industrialisation gave them the means and the motivation to do so until near extinction.
Most of humanity survived on agriculture and sometimes hunting-gathering for last 10k years. People that survived on hunting whales is minuscule. Comparing those two is nonsensical.
Read Moby Dick some time my friend.
The industrial revolution is generally understood to have started somewhere around 1760, Moby Dick took place in approximately 1830, about 10 years before what some historians mark as the end of the agrarian to Industrial shift that is generally termed the Industrial revolution
https://en.wikipedia.org/wiki/Industrial_Revolution
I get sort of wishy-washy from 1830 on, because lots of people put the end of the Industrial revolution as being 1900, but 1840 is a defensible and commonly held position.
the Stepchange show went fairly deep on this topic in their first episode (listened to it recently). https://www.stepchange.show/coal-part-i
> This is just wildly incorrect.
from an global perspective it isn't. Some places sure, like Western Europe, who in some cases had completed enclosure, but remember the new world had only been discovered a few hundred years ago at that point.
Just google maps the north part of South America, even today there are large swathes of undeveloped land across it and back then it was considerably less exploited. At that time it would have appeared infinite, especially to the European industrialists.
> remember the new world had only been discovered a few hundred years ago at that point.
By White people*
we're talking about the fucking industrial revolution, of course this defaults to the European perspective. Unless you wanna spit some new bars about Aztec foundries and train lines connecting meso-america in the 19th century, then the point stands. At that time, the world appeared to the industrialists of the industrial revolution to be infinite. Nor had humanity discovered the terrible side effects of fossil fuels on the atmosphere.
Why are you weirdly making this about race?
Sure, of course it's convenient to ignore the native peoples and pretend that prior to the Industrial Revolution the rest of the world outside of Europe was some untapped well of resources that Europeans had a natural right to.
Who might be swept underfoot in this "Information Revolution", I wonder?
Yes, just the other day I saw someone make a comment about write performance in SQLite without considering the plight of the Baltic peoples in the Northern Crusades. It was really convenient of them to do that, fucking typical.
Sure, because working on a database plugin is the same as, for example, working on mass surveillance tech.
This sort of handwashing is exactly why the natives were treated the way they were.
How do you think they're enabling the mass surveillance tech? SQLite got reach bruv.
Your continued erasure of the Baltic people's continues to cut deep into my heart, and your callous candour to their plight, as you discard any chance to mention them, continues to shock me.
Nobody said anything about Europeans having a "natural right". Bad enough to derail a conversation with irrelevant political nitpicking, unforgiveable to use a strawman to do so. Boo.
It's not irrelevant.
GP made a comparison between what we're going through and the Industrial Revolution. Ignoring the negatives of that revolution - like by acting as though the "new world" was uninhabited/unused and so Europeans had a right to its resources - seems like a bad idea.
"but I can't help but see parallels between today and the Industrial Revolution"
You're not the only one.
The current Pope Leo XIV explicitly named himself after the the previous Leo, Pope Leo XIII, who was pope during the Industrial Revolution (1878-1903) and issued the influential Encyclical Rerum novarum (Rights and Duties of Capital and Labor) in response to the upheaval.
“Pope Leo XIII, with the historic Encyclical Rerum novarum, addressed the social question in the context of the first great industrial revolution,” Pope Leo recalled. “Today, the Church offers to all her treasure of social teaching in response to another industrial revolution and the developments of artificial intelligence.” A name, then, not only rooted in tradition, but one that looks firmly ahead to the challenges of a rapidly changing world and the perennial call to protect those most vulnerable within it.”
https://www.vatican.va/content/leo-xiii/en/encyclicals/docum...
https://www.vaticannews.va/en/pope/news/2025-05/pope-leo-xiv...
[flagged]
As you know, I deeply respect you. Not trying to argue here, just provide my own perspective:
> Why would a writer put an article online if ChatGPT will slurp it up and regurgitate it back to users without anyone ever even finding the original article?
I write things for two main reasons: I feel like I have to. I need to create things. On some level, I would write stuff down even if nobody reads it (and I do do that already, with private things.) But secondly, to get my ideas out there and try to change the world. To improve our collective understanding of things.
A lot of people read things, it changes their life, and their life is better. They may not even remember where they read these things. They don't produce citations all of the time. That's totally fine, and normal. I don't see LLMs as being any different. If I write an article about making code better, and ChatGPT trains on it, and someone, somewhere, needs help, and ChatGPT helps them? Win, as far as I'm concerned. Even if I never know that it's happened. I already do not hear from every single person who reads my writing.
I don't mean that thinks that everyone has to share my perspective. It's just my own.
Agreed, totally! I still write and put stuff online.
But it definitely feels different now. It used to feel like I was tending a public garden filled with other people who might enjoy it. It still kind of feels like that, but there are a handful of giant combine machines grinding their way around the garden harvesting stuff and making billionaires richer at the same time.
It's not enough to dissuade me from contributing to the public sphere, but the vibe is definitely different.
Honestly, it reminds me a lot about the early days of Amazon. It's hard to remember how optimistic the world felt back then, but I remember a time when writing reviews felt like a public good because you were helping other people find good products. It was like we all wanted honest product information and Amazon provided a neutral venue for us to build it. Like Wikipedia for stuff.
But as Amazon got bigger and bigger and the externalities more apparent, it felt less like we were helping each other and more like we were help Bezos buy yet another yacht or media empire. And as the reviews got more and more gamed by shady companies, they became less of a useful public good. The whole commons collapsed.
I worry that the larger web and digital knowledge environment is going that way.
I still intend to create and share my stuff with the world because that's who I want to be. But I'll always miss the early days of the web where it felt like a healthier environment to be that kind of person in.
> But as Amazon got bigger and bigger and the externalities more apparent, it felt less like we were helping each other and more like we were help Bezos buy yet another yacht or media empire.
The Internet-circulating quote comes to mind: Planet Earth is pretty much a vacation resort for around 500 rich people, and the remaining 8 billion of us are just their staff. The Relative Few have got the system set up perfectly so that whatever we do, we're probably serving/enriching them. AI doesn't really change this, but it does further it.
> The Internet-circulating quote comes to mind: Planet Earth is pretty much a vacation resort for around 500 rich people, and the remaining 8 billion of us are just their staff. The Relative Few have got the system set up perfectly so that whatever we do, we're probably serving/enriching them. AI doesn't really change this, but it does further it.
I don't necessarily disagree with the analysis on how Planet Earth is currently setup to be, but something that I've been thinking about lately, is that to the extent we can consume the public image of some of the Relative Few, they seem oddly unhappy.
I think you're right.
Anyone who finds themselves with $100m in their bank account and thinks, "No, I need more," is a person with a hole inside them that can never be filled.
There were a couple of threads on HN this week. "Do you have any unusual hobbies" and "How do you relax". I enjoyed them and was thinking of contributing. Then it occurred to me that my comments would be gold for targeting advertising at me. That is the distrust that has been bred by the data harvesters.
It does feel like the collaborative, free open nature of the web has gone and the optimism that brought… it feels like no one would build Foursquare today. But then I wonder if I’m just old an jaded and to the younger generation creating content, for them the web is open and expressive- just in a different way
I still use swarm every day, and get teased for it all the time.
"So Steve, you're a millennial. What does it mean to 'be the mayor' of something?"
I can totally see that, for sure. I was much more likely to write a review long ago, now I don't even bother. (For buying stuff online, at least.) Maybe I lost my innocence about this stuff a long time ago, and so it's not so much LLMs that broke it for me, but maybe... I dunno, the downfall of Web 2.0 and the death of RSS? I do think that the old internet, for some definition of "old," felt different. For sure. I'll have to chew on this. I certainly felt some shock on the IP questions when all of this came up. I'm from the "information wants to be free" sort of persuasion, and now that largely makes me feel kinda old.
Also I'm not a fan of billionaires, obviously, but I think that given I've worked on open source and tools for so long, I kinda had to accept that stuff I make was going to be used towards ends I didn't approve of. Something about that is in here too, I think.
(Also, I didn't say this in the first comment, but I'm gonna be thinking about the industrial revolution thing a lot, I think you're on to something there. Scale meaningfully changes things.)
I feel the future includes the sentiments you describe. It was a little before my time professionally, but I grew up reading that kind of thinking.
I do think that the open web stuff, decentralized, or at least more decentralized than currently, is the path forward. I've been reading about the AT protocol and it recently becoming an official working group with the IETF.
I feel a second order effect of making decentralized social networking easier, is making individuals more empowered to separate from what they don't believe in. The third order effect is then building separate infrastructure entirely.
As sad as that can be - in my personal opinion it runs the risk of ending the "world wide" part of the web - it appears to be the only way society can avoid enriching the few beyond reason.
> I'm from the "information wants to be free" sort of persuasion, and now that largely makes me feel kinda old.
Me too, 100%. But that was during a moment in time when that information was more likely to be enabling a person who otherwise didn't have as many resources than enabling a billionaire to make their torment nexus 0.1% more powerful.
> I kinda had to accept that stuff I make was going to be used towards ends I didn't approve of. Something about that is in here too, I think.
Yeah, I've mostly made peace with that too.
The way I think about it is that when I make some digital thing and share it with the world, I'm (hopefully!) adding value to a bunch of people. I'm happiest if the distribution of that value lifts up people on the bottom end more than people on the top. I think inequality is one of the biggest problems in the world today and I aspire to have the web and the stuff I make chip away at it.
If my stuff ends up helping the rich and poor equally and doesn't really effect inequality one way or the other, I guess it's fine.
But in a world with AI, I worry that anything I put out there increases inequality and that gives me the heebie-jeebies. Maybe that's just the way things are now and I have to accept it.
> But in a world with AI, I worry that anything I put out there increases inequality and that gives me the heebie-jeebies. Maybe that's just the way things are now and I have to accept it.
This observation doesn't really clash with "information wants to be free." You just have to include LLMs in the category or "information," like Free Software types already do for all software. You don't need to abandon your principles, you should shift your demands. A handful of companies can't be allowed to benefit from free information and then put what they make behind a wall.
> Free Software types already do for all software
Free Software types also create software...they didn't just argue for a better license and try to regulate Sun/others to re-license their software; they wrote free (libre) versions of proprietary software and released it for free (cost), which is what counteracted the "[putting] what they make behind a wall". If you're saying "[some] LLMs should be free", I agree.
I don’t disagree with you, but this has been going on for a while… Google monetized the the by indexing it and monetized what you wanted to find. Facebook monetized the eyeballs from the pictures and posts you added. Now LLMs will monetize all web content. To play devil’s advocate - LLMs do give something back. Those with ideas and no coding experience can now build entire businesses for little to zero cost. This seems different
> A handful of companies can't be allowed to benefit from free information and then put what they make behind a wall.
What is there to prevent them?
Nothing today; but in a democracy, we have the power to make it possible, if people vote the right way.
> the "information wants to be free" sort of persuasion
That was always a luxury of its peculiar historical moment, though, wasn't it? Barlow didn't have to care who paid for the infrastructure, but he was just bloviating.
No, it's as true now as it was then. The intellectual property team didn't win on the merits or by law enforcement; it was the convenience of streaming anything at will for a monthly fee that did the trick.
> it was the convenience of streaming anything at will for a monthly fee that did the trick
That's not the whole story, though. There have been many community-driven projects to bring convenient access to copyrighted works to the masses in a convenient way. You may recall the meteoric success of Popcorn Time. Law enforcement shut them down. Without the hand of the state beating down any popular alternative to legal distribution it absolutely would be the dominant mode of media consumption.
> It used to feel like I was tending a public garden filled with other people who might enjoy it. It still kind of feels like that, but there are a handful of giant combine machines grinding their way around the garden harvesting stuff and making billionaires richer at the same time.
An underrated upside to being harvested is that your voice has now effectively voted in the formation of the machine's constitution. In a broader ecological sense, you've still tended to a public garden, but in this case your work is part of the nutrient base for a different thing.
Broader still: after the machines squeeze all of our inputs into an opaque crystal, that crystal's very purpose is to leak it all back out in measured doses. Yes, "some billionaire" will own the lion's share of that process, but time so far is telling that efforts can be made to distill strong, open, public versions of the same.
> time so far is telling that efforts can be made to distill strong, open, public versions of the same.
I do really hope that part of the longer-term answer for AI is LLMs being run locally.
If raw resources (tree cutting) and manufacturing (book binding) is saturated, a fully-realized economy has just one step left: financialization.
You have to start finding ways to keep people hooked on books and make it a part of their regular lifestyle. One book can't be enough, and after a while you have to convince them to replace the books they already bought. New editions, Author's Footnotes, limited run release, all of the stops have to be pulled out to get consumers to show up en-masse. Because that's what they are - consumers, not readers - wallets to be squeezed until they're bled of all the trust they had in media.
I think about the publications I liked reading as a kid, like Joystiq and Polygon. Some of the best games journalism the industry produced, but inevitably doomed to fail as their competitors monetized further. The rest of traditional media has followed the same path, converging on some mercurial social network marketing tactic as the placeholder for big-picture brand strategy.
> A lot of people read things, it changes their life, and their life is better. They may not even remember where they read these things. They don't produce citations all of the time. That's totally fine, and normal. I don't see LLMs as being any different. If I write an article about making code better, and ChatGPT trains on it, and someone, somewhere, needs help, and ChatGPT helps them? Win, as far as I'm concerned. Even if I never know that it's happened. I already do not hear from every single person who reads my writing.
Not a contradiction but an addendum: plenty of creative pursuits are not about functional value, or at least not primarily. If somebody writes a seemingly genuine blog post about their family trauma, and I as the reader find out it's made-up bullshit, that's abhorrent to me, whether or not AI is involved. And I think it would be perfectly fair for writers who do create similar but genuine content to find it abhorrent that they must compete with genAI, that genAI will slurp up their words, and that genAI's mere existence casts doubt on their own authenticity. It's not about money or social utility, it's about human connection.
The consent question gets weirder when agents have persistent memory. I run agents that accumulate context over weeks — beliefs extracted from observations, relationships with other agents. At what point does an agent's memory become its own work product vs. derivative of its training? There's no legal framework for that.
That seems fine if you're not publishing content for a living. A lot of people are.
> I don't mean that thinks that everyone has to share my perspective. It's just my own.
I think you are walking all around the word "consent" and trying very hard to avoid it altogether.
Your perspective, because it refuses to include any sort of consent, is invalid. No perspective that refuses consent can be valid.
Consent is absolutely important, but that does not mean that every single thing in the entire world requires explicit consent. You did not ask me for consent to use my words in your comment. That does not mean you're a bad person.
Free use is an important part of intellectual property law. If it did not exist, the powerful could, for example, stifle public criticism by declaring that they do not consent to you using their words or likeness. The ability to do that is important for society. It is also just generally important for creating works inspired by others, which is virtually every work. There has to be lines for cases where requiring attribution is required, and cases where it is not.
> You did not ask me for consent to use my words in your comment.
I am not representing your words as mine. I am not using your words to profit off. I am not making a gain by attributing your words to you.
> There has to be lines for cases where requiring attribution is required, and cases where it is not.
You are blurring the lines between "using a quote or likeness" and "giving credit to". I am skeptical that you don't know the difference between the two.
Regardless, any "perspective" that disregards the need to acquire consent is invalid. Even if you are going to ignore it, you have to acknowledge that you don't feel you need any consent from the people you are taking from.
This whole "silence is consent" attitude is baffling.
You made an incredibly strong statement that is much broader than what we are talking about. I am pointing out various cases where I think that broadness is incorrect, I am not equating the two.
I do not think that, if you read, say, https://steveklabnik.com/writing/when-should-i-use-string-vs... , and then later, a friend asks you "hey, should I use String or &str here?" that you need my consent to go "at the start, just use String" instead of "at the start, just use String, like Steve Klabnik says in https://steveklabnik.com/writing/when-should-i-use-string-vs... ". And if they say "hey that's a great idea, thank you" I don't think you're a bad person if you say "you're welcome" without "you should really be saying welcome to Steve Klabnik."
It is of course nice if you happen to do so, but I think framing it as a consent issue is the wrong way to think about it.
We recognize that this is different than simply publishing the exact contents of the blog post on your blog and calling it yours, because it is! To me, an LLM is a transformative derivative work, not an exact copy. Because my words are not in there, they are not being copied.
But again, I am not telling anyone else that they must agree with me. Simply stating my own relationship with my own creative output.
Just wanted to compliment you on your classy attitude and style, along with your solid points. It’s not easy to take that side of the debate. Cheers.
he doesn't have solid points, he conflates fair use with free use (?), ignores thousands of years of attribution history, and equates normal human to human learning with corporate LLMs training on original content (without consent). Great presentation, like you said, to cover the logical defects.
I did say "free use" instead of "fair use," yeah. That's my mistake, thank you for the correction. If I could edit my original comment, I would, mea culpa. Typos happen.
I see. I must congratulate you on your rhetorical prowess, it's nice seeing a professional at work.
Fair use of training data hasn’t yet been settled in court. People here are treating it like it has been. But no amount of wishful thinking or moral arguments will change a verdict saying it’s fine for training data to be used as it has been.
Until that question is settled, it’s disingenuous to dismiss his points out of hand as conflating fair use or ignoring consent.
Even beyond that, the initial legal opinion we do have did in fact point to training being fair use: https://www.reuters.com/legal/litigation/anthropic-wins-key-...
However, I don't feel comfortable suggesting that this is settled just yet, one district judge's opinion does not mean that other future cases may disagree, or we may at some point get explicit legislation one way or the other.
I think the court dropped the ball here. On the one hand, I think they were right that using existing works--copyrighted or otherwise--to train a model was transformable fair use. On the other hand, Anthropic and others trained their models on illicit copies of the works; they (more often than not) didn't pay the copyright holders.
There's a doctrine in Fifth Amendment law called "fruit of the poisonous tree." The general rule is that prosecutors don't get to present evidence in a criminal trial that they gained unlawfully. It's excluded. The jury never gets to see it even if it provides incontrovertible evidence of guilt. The point is to discourage law enforcement from violating the rights of the accused during the investigative process, and to obtain a warrant as the Amendment requires.
It seems to me that the same logic ought to be applied to these companies. They want to make money by building the best models they can. That's fine! They should be able to use all the source data they can legitimately obtain to feed their training process. But if they refuse to do so and resort to piracy, they mustn't be allowed to claim that they then used it fairly in the transformative process.
I was just enumerating some of the issues with the '''solid''' points OP made. Actually addressing them would take too long and be exercise in futility, here, in HN, in april 2026. Why would I put in the effort, for my comment to be flagged and sent to the void? or worse, persisted for ever and used for training without my consent?
And yes, you are right, the legal and moral question of fair use in training data hasn't been settled yet; we agree here.
> But again, I am not telling anyone else that they must agree with me. Simply stating my own relationship with my own creative output.
Look, I'm not saying that you are doing that, I'm pointing out that "Silence is consent" is not as strong an argument that many think it is.
> you don't feel you need any consent from the people you are taking from
In most cases, no, I (and it seems most others) don't feel the need for that, it is only you who seems to have an ideological hangup over this.
> you don't feel you need any consent from the people you are taking from.
What has been "taken", exactly?
refuse consent?
You may need to clarify that thought.
I don't think the poster has a viewpoint that 'refuses consent', their viewpoint is their writing they put for others to view is for others to view, regardless of how it is viewed. They seem to be giving consent, not refusing it, no?
> refuse consent?
Who said anything about refusing consent?
> I think you are walking all around the word "consent" and trying very hard to avoid it altogether.
> Your perspective, because it refuses to include any sort of consent, is invalid. No perspective that refuses consent can be valid.
This is what I was responding to. I do not understand your thinking in this post.
Prior to Industrial Revolution, nobody could go hunt in the woods, because the woods were King’s, and poaching King’s game carried death penalty. Situation was similar on the continent: the tiny slivers of remaining wood lands were off limits.
Granted, things were different in the New World, as a result of mass depopulation event following the Columbian exchange. But even there, the megafauna was hunted to extinction soon after the humans first appeared there.
Anyway, the point is that no, prior to Industrial Revolution, the world was of full of scarcity, not abundance.
I'm halfway through Foundation on Apple TV and this piece landed hard (you had me at Asimov) because of it. Asimov's whole deal with psychohistory is that you can predict what large populations do even when individuals are unpredictable. Seldon doesn't need anyone to be honest; he needs the math to converge on something real about how people actually behave.
LLMs are sort of the inverse of that. They produce text that looks like the statistical aggregate of human knowledge, but nothing underneath is converging on truth. Seldon's math worked because it modeled actual dynamics. LLMs work because they model plausible text. The "jagged competence frontier" Kingsbury describes: crushing multivariable calculus, failing a word problem, is exactly what you'd get from a system that learned the shape of correct answers without learning what makes them correct.
The part of Foundation that feels prescient right now isn't the predicting-the-future stuff. It's the part where everyone can see Empire is hollowing out and the response is to just...keep going. More spectacle, more confidence, less substance holding any of it up. Hmmm, wait...
> Prior to the industrial revolution, the natural world was nearly infinitely abundant.
The opposite is true. Central Europe was almost devoid of trees. Food was scarce as arable land bore little fruit without fertiliser.
Society was Malthusian until the Industrial Revolution.
Can we interpret "abundant" in a Darwinian sense e.g. diversity of life? I would think the industrial farming revolution decreased crop variety over time same for animal lineages aside from the rapid increase in mixed poodle breeds.
To add, I don’t think my ancestor Spaniards for example needed the help of machines to deplete mines in America. They also came already equipped with all kinds of legal systems, including the Requerimiento, which they read out loud to natives in preposterous spectacle.
In general the transition from feudalism to capitalism, including the formation of the legal systems that supported the latter, happened gradually for maybe up to four or five centuries before the steam engine had been invented.
Sure, the Industrial Revolution further accelerated the development of property rights, mercantile, and civil laws, but all in all I don’t think there’s much truth that machines were the primary cause of such developments.
Not really Malthusian. Agricultural societies had adapted to keep the population stable during normal times and bounce back in a generation or two after bad times. Those cultural adaptations stopped working when childhood mortality declined.
Useful land was a scarce resource in more civilized regions, while labor was cheap. Given enough land, subsistence farmers could easily feed themselves outside particularly bad years. But much of the land belonged to local elites, and commoners had to work that land to fund the pursuits of the elites.
If I'm being honest, I've never related to that notion of remuneration and credit being the primary reason to write something. I don't claim to be some great writer or anything, but I do have a blog I write quite often on (though I'm traveling in my wife's Taiwan now and haven't updated it in a while). But for me, I write because it feels good to do so. Sometimes there's a group utility in things like I edit a Google Maps listing to be correct even though "a faceless corporation is going to hoover up my work and profit off it without paying me for my work" and I might pick up a Lime bike someone's dropped into the sidewalk even though "a faceless corporation is externalizing the work of organizing the proper storage of their property on public land without paying the workers" or so on.
I just think it's nice to contribute to the human commons and it's fine if some subset of my fellow organism uses it in whatever way. Realistically, the fact that Brewster Kahle is paid whatever few hundred thousand he's paid for managing a non-profit that only exists because it aggregates other people's work isn't a problem for me. Or that Larry Page and Sergey Brin became ultra-rich around providing a search interface into other people's work. Or that Sam Altman and Dario Amodei did the same through a different interface.
This particular notion doesn't seem to be a post-AI trend. It seems to have happened prior to the big GPTs coming out where people started doing a lot of this accounting for contribution stuff. One day it'll be interesting to read why it started happening because I don't recall it from the past. Perhaps I just wasn't super plugged in to the communities that were complaining about Red Hat, Inc.
It's not that I don't understand if I sold my Subaru to a guy who immediately managed to sell it to another guy for a million times the money. I get that. I'd feel cheated. But if I contributed a little to it, like I did so Google would have a site to list for certain keywords so that they could show ads next to it in their search results, I just find it so hard to be like "That's my money you're using. Pay me!".
You do it as a hobby, that's fine. Some people do it for a living. And while they aren't owed a living doing that specific thing, it is going to be a big problem for them if they can't make money at it anymore.
I'm sure plenty of people feel the same way about software. They make software as a hobby and don't care about remuneration or credit. Meanwhile I write software for my day job and losing the ability to make money from it would be devastating.
> Some people do it for a living.
I was going to write, "not for long," which might be true for some. But then I realized there will always be a difference between LLM output and human writing. We don't read blogs because of their facts, we read them because of how the facts are presented and how the author's personality comes through on the page.
EDIT: That said, LLMs are great at faking it, and a lot of amateur writing will be difficult to distinguish from LLM output. So I'm disagreeing with myself a bit.
But we are talking about "slurping up" IP and regurgitating it right? OK. So if I slurp up Mickey Mouse and output Micky Mouse that's an offense. But what if I slurp up a billion images and output some chimera? That's what the LLMs do. And that's what humans do too.
Ah, I see. It’s just straightforward protectionism like dockworkers opposing automation and so on. That I do comprehend, in fact.
I write software too and I may no longer be able to just do it in the old way. Pretty scary world but also exciting. I can’t imagine trying to restrict LLM software writers on that basis but I can comprehend it as simply self-interest.
Fair enough.
It’s about the amount of time available.
Do you make money writing software? I bet you either try to restrict LLM usage or assign your rights to an employer who does. Putting code in the public domain is pretty rare, and extremely rare for paid work.
I allow them to train on my work as described here (for example) https://code.claude.com/docs/en/data-usage
And I do paste code into CC. I’m not super concerned that they’ll see it.
That’s fine by me. It doesn’t require putting code in the public domain which is something else entirely.
I make money off hosted software so in some sense there is writing involved at one end. But I’m not paid by output tokens.
If your code isn't in the public domain, then anything you haven't explicitly allowed them to train on is restricted for them. They've been ignoring that for anything they can actually get their hands on, but it's there.
Maybe it’s a useless distinction but it feels like we’ve gone through overlapping ages of Communication, Data, Processing, and Information.
First we conquered the ability to move matter and transmit signal, greatly shrinking the world. Next was sensor technology, especially the mobilization of it, and our ability to collect more data than we could ever imagine being able to process. Then we started going crazy with data centres and big data and the idea that maybe we can somehow correlate it all if we just process it enough. And now we’re finally turning data into information, building enormous graphs of correlation without even having to manually reason about a lot of it. Before AI, the hard part was figuring out how to go about finding the signal you needed. Now it’s getting easier at an incredible speed.
>Prior to the industrial revolution, the natural world was nearly infinitely abundant. We simply weren't efficient enough to fully exploit it.
The mammoths disagree.
At what point do we look at 'Industrial Society and its Future' and go from "yeah that'll never happen", "ok some parts of it are happening", to ...? I swear tech folks are the most obtuse people on the planet.
I think it's completely normal. Whenever automation comes knocking, people are inclined to think it's going to flatline conveniently before their job is at risk. LLMs can code now? Cool, they can't code well though can they? Oh they can code pretty well now? Cool, coding was never the hard part of SWE anyway, it's [thing we have no reason to think AI can't beat 99% of humans at at some point], etc
I think SWE as a mainstream profession is much nearer to the end than the beginning, I'm curious and quite scared about what becomes of us.
I don't think you understand. Frankly, AI is a failure if all it does is replace coders. AI needs (given its current investment levels) to conquer all forms of knowledge work. This is an example of tech/industry needing to impose itself on society, rather than society needing it.
That's how human progress works. No one can want or need it because they cannot conceptualize wanting it until someone shows that it is possible. Now, many of those wants become needs.
We can absolutely conceptualize what we want or need. I was born in 1980 in NYC. When I was a boy my father took me to a tech conference where they had a demo of ordering TV shows on demand. It was a miracle, to my young mind. Was this what I needed?
Growing up I had a friend group of misfit boys, who discovered h4ck1ng and phr34king. But we also discovered slackware Linux on 3.5" floppies. We also had to discover ASM and compiling the linux kernel in order to do anything with it. Boys with machines. That wasn't what I needed either.
Later on we did have great things with tech. Google made the world searchable in ways Altavista didn't. I remember strapping the original iPod on my arm to go for runs outside. I didn't even need a car for a while investors subsidized my Uber rides to and from the office.
Now, it seems the US is balanced on a precipice. The economy seems to have an incredible amount of money desperate to grow, but to what purpose. In my lifetime, and in my parents, and their parents before them, when the dollar becomes restless the flag goes forth. The dollar follows the flag.
And here we are at war.
You wouldn't have known about a TV had you not seen it. That is what I mean by, people generally can't conceptualize what they want or need until they see it.
Wants and needs are not the same. We are experiencing the difference in real time. AI does not give society a want or need.
My point was not about the difference, it was about the fact that average people cannot conceptualize new ideas until one person or team invents it, then the average person will want or need it.
As for AI, I and many others want it, and some even need it, in certain use cases. Speak for yourself.
I believe the idea that you (or I) might know better than the 'average people' to be incredibly conceited, arrogant, and frankly wrong. It is an attitude that gives you superiority for having achieved nothing.
I'm not sure what you're even talking about, you're putting words and an argument into my mouth which I never said.
Well then I owe you an apology. Perhaps I inferred too much about your point of view and understood too little, which is my own loss. Sorry.
I think your numbers are off. TAM for office workers is ~20T a year, of which SWE compensation is ~3T. So if they can make 3T x 10% X 5 years = 1.5T that covers their current valuations. It's not as insane as you make out, even not taking into account the other high risk areas like legal, accounting etc
Hit the nail on the head with that framing. So many articles are now coming out addressing the anxieties about adoption of a new technology, but we genuinely don’t really need it as a society.
I still wonder if we really needed the iPhone or many other things we’re told is “progress” and innovation in an arrow of time manner. The future is not set in stone and things need not play out in this manner at all. Unlike the iPhone where most were excited by its possibilities (even if they traded precious privacy in the name of convenience), there’s not a clear reason that this version of LLM driven technologies represent significant upsides than downsides.
>This completely unpends the tenuous balance between creators and consumers. Why would a writer put an article online if ChatGPT will slurp it up and regurgitate it back to users without anyone ever even finding the original article? Who will contribute to the digital common when rapacious AI companies are constantly harvesting it? Why would anyone plant seeds on someone else's farm?
I have been thinking about this. I was pretty amendment a few months ago that AI is going to make a lot of thing worse for everyone because of the externalities of the technology (Data Center Creep, lock in of models, ect) and it probably still will. But then someone suggested to me that I use Claude Code to upgrade my SSG site to the new version because I had been sitting on my ass as the years went by, missing deadline by deadline. I just couldn't put my self into gear to upgrade it. It was massively out of date 10 years plus and I knew it was going to be a nightmare to deal with the problems. I probably was making it more harder than it really was in my head.
So I purchase Claude Code pro and the thing upgrade my site pretty well. There were things it missed because I didn't know the problems existed in the first place until the upgrade was complete, but I had a working updated site in less than an hour. If I had done this myself it would have taking me days/weeks.
So at that point I realized something. Its a tool that can handle good amount of tasks I throw at it as long as I am specific. I think the problem with most people is they expect it to respond like a human. Thats not going to happen, IMHO. Maybe some day it will be more than what it is but right now its just a tool. I don't care what anyone says about AGI and the likes. Its not going to happen with the current iteration (the pattern recognition type) We are going to need more than that if we want to simulate a human brain..
The point is. And I know this is not going to be received very well, mostly because this tech is in the hands of people that are gatekeeping it, is that maybe someday we might reach a point where all of humanities knowledge is put into these things and we can use them to better our lives. Maybe at some point we don't need to hold onto or hoard things as if its the only way we can make a living? And instead we can build things just for the sake of creating it and improve humanity in the process? Obviously the commercial model of these things is not great, that is going to have to be dealt with, but I can see a future where we might be able to fix a lot of humanities problems with this technology as more and more good people put it to use for things that help humanity.
> If all of us can go hunting in the woods and yet there is still game to be found, then there's no compelling reason to define and litigate who "owns" those woods.
Property rights don't just protect natural resources, but labor as well. If I cleared out hunting ground in that forest to be the prime spot to catch animals, I would make sure I can use it when I want.
> a small number of people were able to completely deplete parts of the earth
A small number of people seems inaccurate when there's typically many more individuals in the pipelines for these technologies.
> and in return profit off the knowledge over and over again at industrial scale
Not off just that knowledge, there needed to be a model trained on the data of many others to utilize it.
> Why would a writer put an article online if ChatGPT will slurp it up and regurgitate it back to users without anyone ever even finding the original article?
Who's better at writing in this scenario and what are my motivations? If it's ChatGPT and I did it for money, then I would say I should recognize that I can't compete and find something AI can't do. If it's ChatGPT and I write to convey my ideas in an effort to learn regardless of the bestowment of a new perspective on the reader, I'll keep writing.
> Why would anyone plant seeds on someone else's farm?
They wouldn't unless it was their own way to attain food and survive. And if it's not the only way, they can defer to those with optimal methods to get it the cheapest they can in the market.
A couple thoughts…
Mostly, AIs don’t recite back various works. Yes, there a couple of high profile cases where people were able to get an AI to regurgitate pieces of New York Times articles and Harry Potter books, but mostly not. Mostly, it is as if the AI is your friend who read a book and gives you a paraphrase, possibly using a couple sentences verbatim. In other words, it probably falls under a fair use rule.
Secondly, given the modern world, content that doesn’t appear online isn’t consumed much, so creators who are doing it for the money will certainly continue putting content online. Much of that content will be generated by AIs, however.
You're missing the point. This is the crux of munificent's argument IMO (and I've made variations of it as well)
> We have copyright and intellecual property law already, of course, but those were designed presuming a human might try to profit from the intellectual labor of others.
You getting a summary of a copyrighted work from a friend is necessarily limited by the number of friends you have, the amount of time they have to read stuff and talk to you, and so on. Machines (and AIs) don't have any limitations.
Yes, true. But does that really shift the argument much? An AI is like the most well-read book nerd you’ve ever met. The AI has read everything. They still won’t recite Harry Potter for you at full length and reading what the original author wrote is part of the pleasure.
> An AI is like the most well-read book nerd you’ve ever met. The AI has read everything
But no real book nerd has read everything. Current law was designed for the capabilities of humans.
Sure, we could change current law, but I think that only forces an AI company to buy one copy of every book. I don’t think it gives any sort of royalty stream to anyone beyond that. Copyright is literally the right to make copies. Once I have acquired a copy, I can read it, summarize it, transform it, etc. in myriad ways.
You can't make copies though. AI training requires making copies of materials, even if they're purchased.
Does a literal book nerd profit megacorporations when they bring up books to you? While burning through a household worth of energy in the process? Also, I’d like to talk with such book nerd because they’d have opinions on books, potentially if I brought up something I have read we could exchange thoughts about it, they could make recommendations for me based on their complex experiences instead of statistics from Reddit comments. An LLM can do none of those, while also doing the former. It’s a lose-lose.
Also, a book nerd doesn’t take roughly ~all human created text to train to produce meaningful results. It’s just such a misplaced analogy and people have been making it ever since OpenAI announced chatgpt for the first time - why do people think “an LLM is just a human who read a lot”
Megacorporations making profit is not some evil that needs to be stopped. The economy is not zero sum.
> The economy is not zero sum.
This is true.
But it's not always positive sum, either.
> Megacorporations making profit is not some evil that needs to be stopped.
Externalities are a thing. It's not about the profit per se, but about how (a) the making of that profit might negatively impact others, and (b) the deployment of that profit in pursuit of rent-seeking and other antisocial behavior in order to insure its continued existence might also negatively impact others.
Externalities are a thing, but this isn’t exactly dumping toxic waste into a river.
> If all of us can go hunting in the woods and yet there is still game to be found, then there's no compelling reason to define and litigate who "owns" those woods.
https://en.wikipedia.org/wiki/Feudalism
> Why would a writer put an article online if ChatGPT will slurp it up and regurgitate it back to users without anyone ever even finding the original article?
In the brave new world we're creating, people will write specifically for AI. If you can impress models so much that they "regurgitate" your work, then your work has achieved a kind of immortality.
That is straightforwardly not true, land ownership was very well defined and the people who hunted in it without permission were prosecuted.
> We have copyright and intellecual property law already, of course, but those were designed presuming a human might try to profit from the intellectual labor of others. With AI, we're in the industrial era of the digital world. Now a single corporation can train an AI using someone's copyrighted work and in return profit off the knowledge over and over again at industrial scale.
The idea that copyright simply doesn't apply to AI has more to do with AI companies deciding that they're not going to comply with those laws than the design of the laws. Also a very successful lobby against enforcement by positioning AI as a strategic necessity.
It's not possible (or at least extremely hard) to prove that the final weights they come up with resulted from copyright infringement.
Thats why they are evaluated so high on the stock market. Basically the will steal all the value of intellectual property in a semi legal way.
> Prior to the industrial revolution, the natural world was nearly infinitely abundant. We simply weren't efficient enough to fully exploit it. That meant that it was fine for things like property and the commons to be poorly defined. If all of us can go hunting in the woods and yet there is still game to be found, then there's no compelling reason to define and litigate who "owns" those woods.
I mean, medieval Europe (speaking broadly) had pretty well defined property rights wrt hunting. In fact, the forester at the time was thought of as one of the most corrupt jobs, as they'd commonly have side hustles poaching and otherwise illegally extracting resources from the lands they enforced and kept others from utilizing in a similar way. Quis custodiet ipsos custodes?
The natural world was not meaningfully abundant… Way before the industrial revolution land which was once used for opening hunting was closed off by the ruling class. Even before the Industrial Revolution you had a new class of merchant and factory owners who earned riches to buy land and keep the poor from hunting on it. Much of the natural resources out of reach for the majority and only accessible by those with deep pockets
> Why would a writer put an article online if ChatGPT will slurp it up and regurgitate it back to users without anyone ever even finding the original article?
I'm happy to miss all the stuff that was written just for the financial benefit of the author.
> We are truly in the Information Age now, and I suspect a similar thing will play out for the digital realm.
The analogy seems to be backwards though. It would be as if we previously had a scarcity of land and because of that divided it up into private property so markets could maximize crop yield etc. and then someone came up with a way to grow food on asteroids using robots, and that food is only at the 20th percentile of quality but it's far cheaper. Suddenly food becomes much more abundant and the people who had been selling the 20th percentile food for $5 are completely out of the market because the new thing can do that for $0.05, and the people providing the 50th percentile food for $10 are also taking a hit because the price difference between what they're providing and the 20th percentile stuff just doubled.
The existing plantation owners then want to put a stop to this somehow, or find a way to tax it, but arguments like this have a problem:
> Why would a writer put an article online if ChatGPT will slurp it up and regurgitate it back to users without anyone ever even finding the original article?
This was already the status quo as a result of the internet. Newspapers were slowly dying for 20 years before there was ever a ChatGPT, because they had been predicated on the scarcity of printing presses. If you published a story in 1975 it would take 24 hours for relevant competitors to have it in their printed publication and in the meantime it was your exclusive. The customer who wants it today gets it from you. On top of that, there weren't that many competitors covering local news, because how many local outlets are there with a printing press?
Then blogs, Facebook, Reddit and Twitter come and anyone who can set up WordPress can report the news five minutes after you do -- or five hours before, because now everyone has an internet-connected camera in their pocket so the first news of something happening now comes in seconds from whoever happened to be there at the time instead of the next morning after a media company sent a reporter there to cover it.
The biggest problem we have yet to solve from this is how to trust reports from randos. The local paper had a reputation to uphold that you now can't rely on when the first reports are expected to come from people with no previous history of reporting because it's just whoever was there. But that's the same thing AI can't do either -- it's a notorious confabulist.
And it's the media outlets shooting themselves in the foot with this one, because too many of them have gotten far too sloppy in the race to be first or pander to partisans that they're eroding the one advantage they would have been able to keep. Damn fools to erode the public's trust in their ability to get the facts right when it's the one thing people would otherwise still have to get from them in particular.
Our only hope is that AI in the long run is both powerful and benevolent enough to be its own "whistleblower" in cases of misuse.
I struggle so hard with this anthropomorphism of LLMs. At the end of the day it's a statistical gradient descent predictor with a bunch of "shit" bolted on top to try and steer outputs in a specific way.
They don't have the actual concept of "benevolent"... or a concept of anything at all. Based on an input, they regress down a path of "what is the next most probable statistical token to output next" and that's fucking it, with the bolted-on shit manipulating these outputs a bit.
I don't doubt that at some point there will be some other AI leap, but I'm not even sure it'll be built on this foundation.
What really needs to be developed is an actual artificial brain of sorts. Much like an infant learns language from first principals, a real AI would have a phase of continuous growth, creating actual memories and being able to reflect upon them. I daresay context windows are not that.
I'd really like to encourage anyone to pump the brakes a bit on how these things actually work, and what they actually are. There is a reason sama is pivoting away from video, et. al. and into corporate software coding, much like anthropic.
Stuff gets put online when the reader isn't the customer. Someone is paying for a reader to be told certain things. So it's free at the point of reading.
> It really feels like we're in the soot-covered child-coal-miner Dickensian London era of the Information Revolution and shit is gonna get real rocky before our social and legal institutions catch up
The really discouraging part of this is that it feels like our social and legal institutions don't even care if they catch up or not.
Technology is speeding up and the lag time before anything is discussed from a legal standpoint is way, way too long
I've been looking at things from the same lens since 2023. At the same time, the depletion/hoarding bit isn't new. Companies were already doing this with consumer data, LLMs are just finally the factory moment—now that we have all the raw material we finally have a means of automating production using it.
So, in some ways, I also view LLMs as a pivotal and important wake up call. Companies were already taking the data and using it for a variety of other purposes—it was just way less evident to people when they weren't in direct competition with labor, since, under capital, labor is what we sell.
Either an entire new industry needs to form, or it's finally time to move beyond capitalism. Centralized capital ends up killing itself, because it effectively shuts down its own engine if it kills off consumers, who can only exist in the first place if the wage labor structure holds.
>Prior to the industrial revolution, the natural world was nearly infinitely abundant.
>We had to invent giant legal systems in order to determine who has the right to do that and who doesn't.
Excuse me? The industrial revolution was like 300 years ago. We had laws before that.
> 2017’s Attention is All You Need was groundbreaking and paved the way for ChatGPT et al. Since then ML researchers have been trying to come up with new architectures, and companies have thrown gazillions of dollars at smart people to play around and see if they can make a better kind of model. However, these more sophisticated architectures don’t seem to perform as well as Throwing More Parameters At The Problem. Perhaps this is a variant of the Bitter Lesson.
This is not true and unfortunately this significantly reduced the credibility of this article for me. Raw parameter counts stopped increasing almost 5 years ago, and modern models rely on sophisticated architectures like mixture-of-experts, multi-head latent attention, hybrid Mamba/Gated linear attention layers, sparse attention for long context lengths, etc. Training is also vastly more sophisticated.
The Bitter Lesson is misunderstood. It doesn't say "algorithms are pointless, just throw more compute at the problem", it says that general algorithms that scale with more compute are better than algorithms that try to directly encode human understanding. It says nothing about spending time optimising algorithms to scale better for the same compute, and attention algorithms and LLMs in general have significantly advanced beyond "moar parameters" since the time of Attention is All You Need/GPT2/GPT3.
Literally the paragraph right before the one you quote is this:
> I am generally outside the ML field, but I do talk with people in the field. One of the things they tell me is that we don’t really know why transformer models have been so successful, or how to make them better. This is my summary of discussions-over-drinks; take it with many grains of salt. I am certain that People in The Comments will drop a gazillion papers to tell you why this is wrong.
As I understand it, this article is basically a conglomeration of several attempts at an article that the author has attempted to make over the past decade or so considering the impacts of AI on society. In their own words:
> Some of these ideas felt prescient in the 2010s and are now obvious. Others may be more novel, or not yet widely-heard. Some predictions will pan out, but others are wild speculation. I hope that regardless of your background or feelings on the current generation of ML systems, you find something interesting to think about.
As for the "Bitter Lesson" part, they pretty much directly said that it wasn't the Bitter Lesson exactly, saying it might be a variant of it. Honestly, it felt more like a way of throwing in a reference to something that also might provoke thought, which was done throughout the piece (which again, is the entire point).
It's totally valid to say "this article didn't provoke much thought for me". I'm a bit confused at why you think a lack of specific domain knowledge in a domain that they literally state they are not an expert in would be disqualifying for that purpose though.
The title of the article is “The Future of Everything is Lies, I Guess” and the first part is literally complaining about LLMs being bullshit machines, while the author proceeds to tell confabulations (or lies) of his own. Is there not a bit of irony in that?
If you’re a non-expert in a field, I don’t think it’s a good sign if you’re writing a 10 part article about that field’s impact on society and getting basic facts wrong. How can I trust that the conclusions will be any more credible?
Might the conclusions be correct even if some of the facts are not? Even a stopped clock is right twice a day. And, "approximately correct" is still sometimes valuable.
>Raw parameter counts stopped increasing almost 5 years ago, and modern models rely on sophisticated architectures like mixture-of-experts, multi-head latent attention, hybrid Mamba/Gated linear attention layers, sparse attention for long context lengths, etc.
Agree, I recently updated our office's little AI server to use Qwen 3.5 instead of Qwen 3 and the capability has considerably increased, even though the new model has fewer parameters (32b => 27b)
Yesterday I spent some time investigating it:
- Gated DeltaNet (invented in 2024 I think) in Qwen3.5 saves memory for the KV kache so we can afford larger quants
- larger quants => more accurate
- I updated the inference engine to have TurboQuant's KV rotations (2026) => 8-bit KV cache is more accurate
- smaller KV cache requirements => larger contexts
Before, Qwen3 on this humble infra could not properly function in OpenCode at all (wrong tool calls, generally dumb, small context), now Qwen 3.5 can solve 90% problems I throw at it.
All that thanks to algorithmic/architectural innovations while actually decreasing the parameter count.
5 years ago was the beginning of 2021, just under a year after GPT3 was released (which was not good at doing anything useful). And that model was 175B params.
GPT4 has been widely rumored to have 1.8 trillion params, which is 10x more, and was released 2 years after this "5 years ago" date that you are using here.
So, to quote yourself here, "This is not true and unfortunately this significantly reduced the credibility of this article for me" /s/article/comment
In late 2021, GLaM had 1.2T parameters. It's difficult to find much use of it in the wild and while the benchmarks it uses are rather outdated, it has a HellaSwag score of 76.6% and WinoGrande of 73.5%. GPT3 had 64.3% and 70.2%.
Meanwhile, Gemma 2 9B, a model from July 2024 with 133x fewer parameters than GLaM, scores 82% and 80.6%. Hellaswag and WinoGrande aren't used in modern benchmarks, probably because they're too easy and largely memorised at this point.
And GPT-4 had 1.8T parameters sure, but it's noticeably worse than any modern model a fraction of the size, and the original incarnation was ridiculously expensive per token. And in any case, its number of parameters was only possible due using mixture-of-experts, which I would definitely classify as a sophisticated architecture as opposed to just throwing more parameters at a vanilla transformer. Even in 2021 GLaM was a MoE because the limits of scaling dense transformers had already been hit.
MoE has made it vastly easier to increase total parameters (and recent open models are really quite large) but it's also hard to compare a MoE with an earlier dense model.
> hybrid Mamba/Gated linear attention layers,
Do any large-scale architectures use mamba? I was under the impression that people don't use it yet due to lack of efficient implementations.
> Training is also vastly more sophisticated
Is it? In what ways?
Qwen3.5 uses Gated Delta Networks which is essentially Mamba 2 + Delta Rule. It’s quite hardware efficient.
> Is it? In what ways?
Just the reinforcement learning for reasoning, and then tool use for agents, could be its own topic.
Yeah I also came here to be one of those People In The Comments the author refers to.
Transformers are not magical. They are just a huge improvement over other architectures at the time such as LSTMs and RNNs and even CNNs. They allowed us to throw more and more compute at the problem of next token prediction. And we’ve been riding that horse ever since.
Another big advancement that deserves mentioning is “reasoning” models that have the opportunity to spit out thinking tokens before giving a final answer.
None of this is to say transformers are the most principled approach. But they work.
This sounds almost identical to the article that's literally linked at the end of the paragraph that the parent comment quoted: https://www.cs.utexas.edu/~eunsol/courses/data/bitter_lesson...
I don't think anything you're saying here is in disagreement with the points they're making.
Transformers' greatest improvement over RNN/LSTM was to enable better parallelization of large-scale training. This is what enabled language models to become "large". But when controlling for overall size, more RNN/LSTM-like approaches seem to be more efficient, as seen e.g. in state space models. The transformer architecture does add some notable capabilities in accounting for long-range dependencies and "needle in a haystack" scenarios, but these are not a silver bullet; they matter in very specific circumstances.
With modern training techniques, RNNs (not just linear SSMs, potentially even vanilla LSTMs) can scale just as well as transformers or even better when it comes to enormous context lengths. Dot-product attention has better performance in a number of domains however (especially for exact retrieval) so the best architectures are likely to remain hybrid for now.
>With modern training techniques, RNNs (not just linear SSMs, potentially even vanilla LSTMs) can scale just as well as transformers or even better when it comes to enormous context lengths.
That's not true. Modern training techniques aren't enough. Vanilla RNNs with modern training techniques still scale poorly. You have to make some pretty big architectural divergences (throwing away recurrency during training) to get a RNN to scale well. None of the big labs seem to be bothered with hybrid approaches.
> That's not true. Modern training techniques aren't enough. Vanilla RNNs with modern training techniques still scale poorly. You have to make some pretty big architectural divergences (throwing away recurrency during training) to get a RNN to scale well.
SSMs move the non-linearity outside of the recurrence which enables parallelisation during training. It is trivial to do this architectural change with an LSTM (see the xLSTM paper). Linear RNNs are still RNNs.
But you can still keep the non linearity by training with parallel Newtown methods, which work on vanilla LSTMs and scale to billion of parameters.
> None of the big labs seem to be bothered with hybrid approaches.
Does Alibaba not count? Qwen3.5 models are the top performers in terms of small models as far as my tests and online benchmarks go.
>SSMs move the non-linearity outside of the recurrence which enables parallelisation during training. It is trivial to do this architectural change with an LSTM (see the xLSTM paper). Linear RNNs are still RNNs.
Removing the non-linearity from the recurrence path is exactly what constitutes a "pretty big architectural divergence." A linear RNN is an RNN in a structural sense, certainly, but functionally it strips out the non-linear state transitions that made traditional LSTMs so expressive, entirely to enable associative scans. The inductive bias is fundamentally altered. Calling that simply 'modern training techniques' is disingenous at best.
>But you can still keep the non linearity by training with parallel Newtown methods, which work on vanilla LSTMs and scale to billion of parameters.
That does not scale anywhere near as well as Transformers in compute spend. It's paper/research novelty. Nobody will be doing this for production.
>Does Alibaba not count? Qwen3.5 models are the top performers in terms of small models as far as my tests and online benchmarks go.
I guess there's some misunderstanding here because Qwen is 100% a transformer, not a hybrid RNN/LSTM whatever.
> That does not scale anywhere near as well as Transformers in compute spend. It's paper/research novelty. Nobody will be doing this for production.
What exactly makes you so confident?
The world is not just labs that can afford billion dollar datacentres and selling access to SOTA LLMs at $30/Mtokens. Transformers are highly unsuitable for many applications for a variety of reasons and non-linear RNNs trained via parallel methods are an extremely attractive value proposition and will likely feature in production in the next products I work on.
> I guess there's some misunderstanding here because Qwen is 100% a transformer, not a hybrid RNN/LSTM whatever.
See the Qwen3.5 Huggingface description: https://huggingface.co/Qwen/Qwen3.5-27B > Efficient Hybrid Architecture: Gated Delta Networks combined with sparse Mixture-of-Experts deliver high-throughput inference with minimal latency and cost overhead.
> It remains unclear whether continuing to throw vast quantities of silicon and ever-bigger corpuses at the current generation of models will lead to human-equivalent capabilities. Massive increases in training costs and parameter count seem to be yielding diminishing returns. Or maybe this effect is illusory. Mysteries!
I’m not even sure whether this is possible. The current corpus used for training includes virtually all known material. If we make it illegal for these companies to use copyrighted content without remuneration, either the task gets very expensive, indeed, or the corpus shrinks. We can certainly make the models larger, with more and more parameters, subject only to silicon’s ability to give us more transistors for RAM density and GPU parallelism. But it honestly feels like, without another “Attention is All You Need” level breakthrough, we’re starting to see the end of the runway.
I see a lot of researchers working on newer ideas so I wouldn't be surprised if we get a breakthrough in 5-10 years. After all, the gap between AlexNet and Attention is All You Need was only 6 years. And then Scaling Laws was about 3-4 years after that. It might seem like not much progress is being made but I think that's in part because AI labs are extremely secretive now when ideas are worth billions (and in the right hands, potentially more).
Of course 5-10 years is a long time to bang our heads against the wall with untenable costs but I don't know if we can solve our way out of that problem.
I think we will see models becoming small reasoning core which don't remember tonnes of facts but can reason with data fed to it or they can search.
The echoes of A.I. winter.
I think in domains like Math and Software Engineering, they are less constrained by training data anyway. They can synthetically generate and validate programs. To what extent that scales into novel insights is a different matter, but I think they dream of the AlphaGo Zero moment at least in verifiable domains.
> I’m not even sure whether this is possible.
Based on what's happened so far, maybe. At least that's exactly how we got to the current iteration back in 2022/2023, quite literally "lets see what happens when we throw an enormous amount data at them while training" worked out up until one point, then post-training seems to have taken over where labs currently differ.
Right, but we played the scaling card and it worked but is now reaching limits. What is the next card? You can surely argue that we can find a new one at any time. That’s the definition of a breakthrough. I just don’t see one at the moment.
> I just don’t see one at the moment.
Did you see the one before the current one was even found? Things tend to look easy in hindsight, and borderline impossible trying to look forward. Otherwise it sounds like you're in the same spot as before :)
That’s what I’m said. Breakthroughs happen. No doubt about it, and they are unpredictable. Hence a breakthrough. But right now we’re using up runway with nothing yet identified to take us to the next level. And while sometimes breakthroughs happen, sometimes they don’t.
better tooling and integration
We pay people to create more high quality tokens (mercor, turing) which are then fed into data generating processes (synthetic data) to create even more tokens to train on
But does that really help, or do you get distortion? The frequency distribution of human generated content moves slowly over time as new subjects are discussed. What frequency distribution do those “data generating processes” use? And at root, aren’t those “data generating processes” basically just another LLM (I.e., generating tokens according to a probability distribution)? Thus, aren’t we just sort of feeding AI slop into the next training run and humoring ourselves by renaming the slop as “synthetic data?” Not trying to be argumentative. I’m far from being an AI expert, so maybe I’m missing it. Feel free to explain why I’m wrong.
That's the problem in a nutshell. There is an art to how you generate the synthdata so that you don't get crappy trained models (especially when mistakes cost XX million dollars).
It's also theoretically why facebook paid 14bn for alex wang and scale ai
> The current corpus used for training includes virtually all known material.
This is just totally incorrect. It's one of those things everyone just assumes, but there's an immense amount of known material that isn't even digitized, much less in the hands of tech companies.
What large caches of undigitized content exists? Surely, not everything has been digitized, but I can’t think it’s much in percentage terms.
The amount of private data that is locked up inside private internal databases is huge. This is especially true of regulated industries. There is a wealth of data - financial data showing how to budget for things, pricing data on various products that are B2B, standard operating procedures at mature companies that have gone through various revisions, designs for manufacturing plants so people don't keep reinventing and making the same mistakes again, and on and on.
The Vatican Library contains roughly 1.1 million printed books and around 75,000 codices, only a small percentage of which have been digitised.
Reddit alone contains about the same quantity of text (~10 billion posts * 10 words per post, vs 1 million books * 100k words per book). Messaging and document platforms (google docs, slack, discord, telegram, etc.) probably each have 1-3 orders of magnitude more than reddit. To your/GP's point though, those private platforms probably haven't been slurped up by LLMs yet.
Which is what percent of the world’s content? 0.000000001% or something similar. It’s nothing in the scheme of things. To put it another way, if we were to digitize that continent and train on it, our AIs would not get noticeably better in any way. It doesn’t move the needle.
1.1 million being 0.000000001% implies a total count of 1e17 books in the world - the real number is closer to 1e8.
You’re missing the point. And we’re not just talking about books, whatever that might mean. We’re talking about all documents ever made. Every magazine article, every blog and web page, every Word doc, etc. I’m pretty sure that whatever is in the Vatican archives is tiny by comparison. Given the age of the Vatican archives, I can also guarantee that many of those “books” are nothing more than page fragments. Very few will be full codices or long scrolls. Many will date before the printing press when document production was slow and laborious.
I think the discussion has to be more nuanced than this. "LLMs still can't do X so it's an idiot" is a bad line of thought. LLMs with harnesses are clearly capable of engaging with logical problems that only need text. LLMs are not there yet with images, but we are improving with UI and access to tools like figma. LLMs are clearly unable to propose new, creative solutions for problems it has never seen before.
> LLMs are clearly unable to propose new, creative solutions for problems it has never seen before.
LLMs are incredibly useful but I'm not sure about this statement.
It is proposing stuff that I haven't seen before, but I don't know about it is new or creative from the entirety of collective human knowledge.
I'm not sure if you misread the statement you quoted or I'm misreading yours, but it doesn't sound like you're really disagreeing with their point. Did you miss the "un" in "unable", or am I misunderstanding you as also saying that you don't consider them to be creative?
> LLMs with harnesses are clearly capable of engaging with logical problems that only need text.
To some extent. It's not clear where specifically the boundaries are, but it seems to fail to approach problems in ways that aren't embedded in the training set. I certainly would not put money on it solving an arbitrary logical problem.
> To some extent. It's not clear where specifically the boundaries are, but it seems to fail to approach problems in ways that aren't embedded in the training set. I certainly would not put money on it solving an arbitrary logical problem.
In what way can you falsify this without having the LLM be omniscient? We have examples of it solving things that are not in the training set - it found vulnerabilities in 25 year old BSD code that was unspotted by humans. It was not a trivial one either.
Here's an odd example of testing, but I design very complex board and card games, and LLMs are terrible at figuring out whether they make sense or really even restating the rules in a different wording.
I thought they would be ideal for the job, until I realized that it would just pretend that the rules worked because they looked like board game rules. The more you ask it to restate, manipulate or simulate the rules, the more you can tell that it's bluffing. It literally thinks every complicated set of rules works perfectly.
> it found vulnerabilities in 25 year old BSD code that was unspotted by humans.
I don't think the age of the code makes the problem more complex. Finding buffers that are too small is not rocket science, bothering to look at some corner of some codebase that you've never paid attention to or seen a problem with is. AI being infinitely useful (cheap) to sic on pieces of codebase nobody ever carefully looks at is a great thing. It's not genius on the part of the AI.
> Here's an odd example of testing, but I design very complex board and card games, and LLMs are terrible at figuring out whether they make sense or really even restating the rules in a different wording.
I'm positive that they are perfectly fine and will a pretty good job. Did you actually try it?
Eh, I can see their point, I think. The models can restate the rules differently, I'm sure, but it sounds like the GP is saying that LLMs can't tell whether the rules are well-balanced.
It would be interesting to see some example problems along those lines. Design some games with complex rules, including one or two of the most subtle game-wrecking bugs you can think of, and ask the models if they can spot them.
In fact that sounds more interesting the more I think about it. Intensive RL on that sort of thing might generalize in... let's say useful ways.
Solving arbitrary logical problems seems to be equivalent to solving the halting problem so you are probably wise not to make that bet.
> LLMs are not there yet with images
https://genai-showdown.specr.net/image-editing
There's been a lot of progress there, it's just that an LLM that's best for, say coding, isn't going to be also the best for image edit.
> "LLMs still can't do X so it's an idiot"
Let’s be careful. That’s a straw man. I don’t know anyone who says that. Aphyr says in the article that AIs can do things. But they have been marketed as “intelligent,” and I agree with Aphyr that the word is suggesting way more than AIs currently deliver. They do not reason and they do not think and are not truly intelligent. As the article says, they are big wads of linear algebra. Sometimes, that’s useful.
> They do not reason
How do you disprove it?
We know that they do not reason because we know the algorithm behind the curtain. The model is generating the next token via model weights and some randomness. That’s all. It not reasoning. Sometimes it has an appearance of reasoning, but not if you know how it works. It doesn’t matter that the model manufacturer marketing department slaps a “Reasoning!” sticker on the side of the model. It’s not actually doing that. As an analogy, sometimes a stage magician in Las Vegas makes it seem that he’s making a woman disappear and a tiger appear in her place, but we all know that’s not what is really happening; It’s just a clever trick.
Well, could you define what reasoning actually means? What would an AI need to do to be considered capable of reasoning? What is the core difference between what we do that is considered reasoning verse what AI currently does that is not considered reasoning?
To be clear, I am not making a statement as to whether AI reasons or not. Its just slippery to say something isn't or can't do X when we can't really define X. Perhaps if we can put it down as an outcome rather than an, in my opinion, currently impossible to accurately define characteristic of a thing.
What is reasoning?
> LLMs with harnesses are clearly capable of engaging with logical problems that only need text.
> LLMs are clearly unable to propose new, creative solutions for problems it has never seen before.
How do you reconcile this with this article that the author linked? It's not a novel problem, and it's only text: https://medium.com/the-generator/one-word-answers-expose-ai-...
I guess it's a form of engagement to give a wildly wrong answer, but I'm not convinced that the extra nuance you've introduced is really all that nuanced either.
Thank you for putting it so succinctly.
I keep explaining to my peers, friends and family that what actually is happening inside an LLM has nothing to do with conscience or agency and that the term AI is just completely overloaded right now.
> I keep explaining to my peers, friends and family that what actually is happening inside an LLM has nothing to do with conscience or agency
What would the insides have to look like to have anything to do with conscience or agency?
One thing that has happened is that "AI" has been an academic discipline since literally the 1950s. The term was originally used in the hope that we would soon be able to emulate human minds. This turned out to be hard, but the name stuck to the discipline.
Now, suddenly, this name has been broadcast to every human in the world more or less. To them, it's a new term, and it obviously means something human mind-like. But to people who work on AI, that's not generally what it means. (Which isn't to say that some of them don't think we're near to achieving that; they just use other terms like "AGI" for that goal). So the name, which has a long history, is deceptive to people who aren't familiar with computer science.
> Now, suddenly, this name has been broadcast to every human in the world more or less. To them, it's a new term, and it obviously means something human mind-like. But to people who work on AI, that's not generally what it means. (Which isn't to say that some of them don't think we're near to achieving that; they just use other terms like "AGI" for that goal). So the name, which has a long history, is deceptive to people who aren't familiar with computer science.
I think it's even worse than that: people were familiar with the term already, but from science fiction, where it referred to actually human-level intelligence. It's similar to the "hoverboard" thing from a while back, except this time with profoundly higher stakes and requires for more technical knowledge to be able to see that it is in fact touching the ground.
AI is exactly the right term: the machines can do "intelligence", and they do so artificially.
Just like we have machines that can do "math", and they do so artificially.
Or "logic", and they do so artificially.
I assume we'll drop the "artificial" part in my lifetime, since there's nothing truly artificial about it (just like math and logic), since it's really just mechanical.
No one cares that transistors can do math or logic, and it shouldn't bother people that transistors can predict next tokens either.
> Just like we have machines that can do "math", and they do so artificially.
Nobody calls calculators "artificial mathemeticians", though; we refer to them by a unique word that defines what they can and can't do in a far less fanciful and ambiguous way.
> AI is exactly the right term: the machines can do "intelligence", and they do so artificially.
AI in pop culture doesn't mean that at all. Most people impression to AI pre-LLM craze was some form of media based on Asmiov laws of robotics. Now, that LLMs have taken over the world, they can define AI as anything they want.
In 2018, ie “pre-LLM”, the label “AI” was already stamped to everything, so I highly doubt that most people thought that their washing machines are sentient in any way. I remember this starkly, because my team was responsible at Ericsson (that time, about 120k employees) for one of the crucial step to have models in production, and basically every single project wanted that stamp.
The shift in meaning has been slowly diluted more and more across decades.
> Most people impression to AI pre-LLM craze was some form of media based on Asmiov laws of robotics.
I'll reveal you a secret: "positronic brains" are just very fast parallel computers running LLMs.
> what actually is happening inside an LLM has nothing to do with conscience or agency
What makes you think natural brains are doing something so different from LLMs?
Structurally a transformer model is so unrelated to the shape of the brain there's no reason to think they'd have many similarities. It's also pretty well established that the brain doesn't do anything resembling wholesale SGD (which to spell it is evidence that it doesn't learn in the same way).
>Structurally a transformer model is so unrelated to the shape of the brain there's no reason to think they'd have many similarities.
Substrate dissimilarities will mask computational similarities. Attention surfaces affinities between nearby tokens; dendrites strengthen and weaken connections to surrounding neurons according to correlations in firing rates. Not all that dissimilar.
Sure the implementation details are different.
I suppose I should have asked by what definition of "consciousness and agency" are today's LLMs (with proper tooling) not meeting?
And if today's models aren't meeting your standard, what makes you think that future LLMs won't get there?
Given the large visible differences in behavior and construction, akin to the difference between a horse and a pickup truck, I would ask the reverse question: In what ways do LLMs meet the definition of having consciousness and agency?
Veering into the realm of conjecture and opinion, I tend to think a 1:1 computer simulation of human cognition is possible, and transformers being computationally universal are thus theoretically capable of running that workload. That being said, that's a bit like looking at a bird in flight and imagining going to the moon: only tangentially related to engineering reality.
> In what ways do LLMs meet the definition of having consciousness and agency?
Agency: an ability to make decisions and act independently. Agentic pipelines are doing this.
Consciousness: something something feedback[1] (or a non-transferable feeling of being conscious, but that is useless for the discussion). Recurrent Processing Theory: A computation is conscious if it involves high-level processed representations being fed back into the low-level processors that generate it.
Tokens are being fed back into the transformer.
> that's a bit like looking at a bird in flight and imagining going to the moon: only tangentially related to engineering reality.
Is it? Vacuum of space is a tangible problem for aerodynamics-based propulsion. Which analogous thing do we have with ML? The scaled-up monkey brain[2] might not qualify as the moon.
[1] https://www.astralcodexten.com/p/the-new-ai-consciousness-pa...
[2] https://www.frontiersin.org/journals/human-neuroscience/arti...
What about modern LLMs isn't "agentic" enough?
Doesn't matter if they're conscious for that. They're clearly capable of goal oriented behavior.
[dead]
If platonic representation hypothesis holds across substrates, then it might matter very little, in the end. It holds across architectures in ML, empirically.
The crowd of "backpropagation and Hebbian learning + predictive coding are two facets of the very same gradient descent" also has a surprisingly good track record so far.
For starters, natural brains have the innate ability to differentiate between things that it knows and things that it have no possibility of knowing...
https://personal.utdallas.edu/~otoole/CGS2301_S09/7_split_br...
See page 53. While it is absolutely more prevelant in LLMs, human brains can also want a story for why their brains do things they are't plugged into.
Lol. Are you sure about that or you just made it up?
Modern LLMs are fairly good at that as well.
But that is bolted on and is not a core behavior.
Does it matter? Evolution is the brain's very own "pre-training". Hundreds of millions of years of priors hardwired.
We can do that for AIs too - pre-train on pure low Kolmogorov complexity synthetics. The AI then "knows things" before it sees any real data. Advantageous sometimes. Hard to pick compute efficient synthetics though.
Any amount of reading into how we understand brains and LLMs to work.
> Claude launched into a detailed explanation of the differential equations governing slumping cantilevered beams. It completely failed to recognize that the snow was entirely supported by the roof, not hanging out over space. No physicist would make this mistake, but LLMs do this sort of thing all the time.
You have to meet some physicist friends of mine then. They are likely to assume that the roof is spherical and frictionless.
To be fair, starting with a toy model to get a first order approximation then building on it is kind of how theoretical science is done.
I think it's too early to declare the Turing test passed. You just need to have a conversation long enough to exhaust the context window. Less than that, since response quality degrades long before you hit hard window limits. Even with compaction.
Neuroplasticity is hard to simulate in a few hundred thousand tokens.
"You're absolutely right!"
I think for a while the test was passed. Then we learned the hallmark characteristics of these models, and now most of us can easily differentiate. That said -- these models are programmed specifically to be more helpful, more articulate, more friendly, and more verbose than people, so that may not be a fair expectation. Even so, I think if you took all of that away, you'd be able to differentiate the two, it just might take longer.
Right. I think the modern LLMs are quite good at mimicking human words, but we were initially taken in like we were in the 1960s by ELIZA. It’s a (increasingly sophisticated) magic trick, but it’s just a trick.
It was not meant as a pass/fail
For as rigorous of a Turing test as you present, I believe many (or even most) humans would also fail it.
How many humans seriously have the attention span to have a million "token" conversation with someone else and get every detail perfect without misremembering a single thing?
Response quality degrades long before you hit a million tokens.
But sure, let's say it doesn't. If you interact with someone day after day, you'll eventually hit a million tokens. Add some audio or images and you will exhaust the context much much faster.
However, I'll grant you that Turing's original imitation game (text only, human typist, five minutes) is probably pretty close, and that's impressive enough to call intelligence (of a sort). Though modern LLMs tend to manifest obvious dead giveaways like "you're absolutely right!"
But context window exhaustion does not look like mere forgetfulness, but more like loss of general coherence, like getting drunk.
Doesn't the Turing test require a human too, to be compared to the AI?
It's weird, I don't know how normally pedantic comp sci. people let this meme that the Turing test is beaten by LLMs to spread so unchallenged. As far as I'm aware, there is no restriction in the Turing test that demands that the interrogator be ignorant of the latest state-of-art in computing (and AI tech), nor is there a strict time limit enforced for the questioning?
Given these conditions, it should be relatively easy for the interrogator to expose the AI in this current day and age.
There are a lot of difference kinds of LLMs. 0 of the ones I've encountered are good writers, in fact all of them are horrible at it.
But I wonder if there's one out there that I don't know about with a different kind of training that actually is good at writing and fun to talk to for a long time. (granted somepeople love talking to gpt 4, but also some people loved talking to ELIZA so clearly some people have a super high tolerance for slop.)
I don't know. Practically, LLMs are already better conversation partners on any topic compared to the average human I have access to. This also holds in reverse, of course - if someone wants me to explain something, usually they'd be better off asking an LLM.
I hesitate to tamper with an internet master's title, but "The Future of Everything is Lies, I Guess" doesn't really summarize what in fact is a balanced, informed overview which (to me at least) is above the median for one of these thought pieces. Since it's also baity and the HN guidelines ask for such titles to be rewritten, I've taken the license.
In such cases we always try to find a phrase from the article itself which expresses what it's saying in a representative way. (There nearly always is one.) In this case, both the very first and very last sentences do this, and it's interesting that they more or less agree. So I plucked the last sentence and put it above.
Edit: oof, I missed that this is actually the first part of a long series. Not sure what we'll do about the others; I expect some of those will make the frontpage as well.
Changing the title was a good call.
The article has a good take on the "lie" problem. We know about the hallucination problem, which remains serious. The "lie" problem mentioned is that if you ask an LLM why it said or did something, it has no information of how it got a result. So it processes the "why" as a new query, and produces a plausible explanation. Since that explanation is created without reference to the internals of how the previous query was processed, it may be totally wrong. That seems to be the type of "lie" the author is worried about in this essay.
(Yes, humans do that too.)
I appreciate the curation you do, dang. I often notice a headline get updated and the result is always a significant improvement.
Honestly, good call on the title. The original one is far less representative. Far better at clickbait though.
Thank you both! I totally missed the sidebar on the OP which explains that this is Part 1 of what will be a long series. Not sure how we'll handle that...
Articles like this should approach topics on consciousness with more humility than is displayed here.
We don’t even agree on a good definition of what’s going on inside our own heads yet, what gives you the confidence to say that what goes on inside an LLM can’t be conscious?
Obviously, the LLMs lack the divine spark, so they can't be conscious. Same as clones, IVF babies, or half of all the twins.
Jest aside, I do agree. If you list out every prominent theory of consciousness, you'd find that about a quarter rules out LLMs, a quarter tentatively rules LLMs in, and what remains is "uncertain about LLMs". And, of course, we don't know which theory of consciousness is correct - or if any of them is.
With LLMs, materialists have got a silicone idol to worship. They believe that they know everything about this idol because they've created it, and at the same time they believe that LLMs have a secret sauce that makes it intelligent. Looking at how they are trying to extract intelligence from it reminds me of alchemists of the past who tried to extract gold from lead.
Even if nothing substantial come out of this, having shortest paths to the corpus of all human expressions in all languages! and media formats is quite something by itself, what could be the ultimate hard information retrieval tool is hiding those trace behind untracked convolutions is the real shame here. Found so much real information in between less and less hallucinations already that was impossible to retrieve otherwise in that time frame. Basically tokenrank kills pagerank.
> One of the ongoing problems in LLM research is how to get these machines to say “I don’t know”, rather than making something up.
To be fair, I've known humans who are like this as well.
If you change it from asking a question to giving an instruction, how many humans do you know that have trouble saying no to things that aren't reasonable? I'd argue that pretty much every human will refuse to do most things you might instruct them to do, whereas an LLM will happily attempt most things you ask them to do for you, regardless of whether they're capable of succeeding, and it's up to you to figure out if they actually did it right or not. There are tasks where this is extremely useful, but they're ones that are extremely low risk and can easily be audited upon completion. This isn't anywhere near the level of what a human is capable of.
This is a limitation of the training data. If you were uncertain about something, you wouldn’t write a book about it. The kinds of people you’re talking about tend to generate far more text in their lives than others, because they can spend more time generating - writing books, blogposts, whatever - and less time thinking and working and actually doing things. The models never say they’re uncertain because we never say we’re uncertain, or at least we don’t write it down anywhere.
Those people aren't the ones doing the work though.
Some people point at LLMs confabulating, as if this wasn’t something humans are already widely known for doing.
I consider it highly plausible that confabulation is inherent to scaling intelligence. In order to run computation on data that due to dimensionality is computationally infeasible, you will most likely need to create a lower dimensional representation and do the computation on that. Collapsing the dimensionality is going to be lossy, which means it will have gaps between what it thinks is the reality and what is.
They key capability that humans have that I've yet to see in an LLM is the ability to recognize when they would not be capable of doing a task well and refuse to do it poorly instead. The only times I've ever seen LLMs give up on a problem are when the prompting is very explicitly crafted to try to elicit a response like that when necessary or after very long back-and-forth exchanges where they get repeated feedback about unsatisfactory results. I think this has pretty dire implications in terms of what the consequences are for deploying them in any scenario where failure has significant risk or the output can't be immediately audited for correctness.
The concern for me about LLMs confabulating is not that humans don't do it. It's that the massive scale at which LLMs will inevitably be deployed makes even the smallest confabulation extremely risky.
I don't understand this. Many small errors distributed across a large deployment sounds a lot like normal mode of error prone humans / cogs / whatevers distributed over a wide deployment.
There's a difference between 1000 diverse humans with varied traits making errors that should cancel out because of the law of large numbers vs 10 AI with the same training data making errors that would likely correlate and compound upon each other.
Let's say a given B2B system deployment typically requires 100 custom behaviours/scripts and 3 years worth of effort. A team of ten people can execute such a deployment in 3-4 months. The team has the capacity to fix up issues caused by small human errors as they arise, since they show up roughly once a week.
With the advent of LLMs, a new deployment now takes 3 days. Consequently, errors requiring human attention crop up several times a day.
I have yet to see a comparison of human vs. LLM confabulation errors at scale.
"Many small errors" makes a presumption about LLM confabulation/hallucination that seems unwarranted. Pre-LLM humans (and our computers) have managed vast nuclear arsenals, bioweapons research, and ubiquitous global transport - as a few examples - without any catastrophic mistakes, so far. What can we reasonably expect as a likely worst case scenario if LLMs replacing all the relevant expertise and execution?
Your project vue-skuilder has 6 github action steps devoted to checking the work you do before it's allowed to go out. You do not trust yourself to get things right 100% of the time.
I am watching people trust LLM-based analysis and actions 100% of the time without checking.
> Some people point at LLMs confabulating, as if this wasn’t something humans are already widely known for doing.
I think we need to start rejecting anthropomorphic statements like this out of hand. They are lazy, typically wrong, and are always delivered as a dismissive defense of LLM failure modes. Anything can be anthropomorphized, and it's always problematic to do so - that's why the word exists.
This rhetorical technique always follows the form of "this LLM behavior can be analogized in terms of some human behavior, thus it follows that LLMs are human-like" which then opens the door to unbounded speculation that draws on arbitrary aspects of human nature and biology to justify technical reasoning.
In this case, you've deliberately conflated a technical term of art (LLM confabulation) with the the concept of human memory confabulation and used that as a foundation to argue that confabulation is thus inherent to intelligence. There is a lot that's wrong with this reasoning, but the most obvious is that it's a massive category error. "Confabulation" in LLMs and "confabulation" in humans have basically nothing in common, they are comparable only in an extremely superficial sense. To then go on to suggest that confabulation might be inherent to intelligence isn't even really a coherent argument because you've created ambiguity in the meaning of the word confabulate.
>this LLM behavior can be analogized in terms of some human behavior, thus it follows that LLMs are human-like
No, the argument is "this behavior is similar enough to human behavior that using it as evidence against <claim regarding LLM capability that humans have> is specious"
>"Confabulation" in LLMs and "confabulation" in humans have basically nothing in common
I don't know why you think this. They seem to have a lot in common. I call it sensible nonsense. Humans are prone to this when self-reflective neural circuits break down. LLMs are characterized by a lack of self-reflective information. When critical input is missing, the algorithm will craft a narrative around the available, but insufficient information resulting in sensible nonsense (e.g. neural disorders such as somatoparaphrenia)
> No, the argument is "this behavior is similar enough to human behavior that using it as evidence against <claim regarding LLM capability that humans have> is specious"
I'm not really following. LLM capabilities are self-evident, comparing them to a human doesn't add any useful information in that context.
> LLMs are characterized by a lack of self-reflective information. When critical input is missing, the algorithm will craft a narrative around the available, but insufficient information resulting in sensible nonsense (e.g. neural disorders such as somatoparaphrenia)
You're just drawing lines between superficial descriptions from disparate concepts that have a metaphorical overlap. It's also wrong. LLMs do not "craft a narrative around available information when critical input is missing", LLM confabulations are statistical, not a consequence of missing information or damage.
>LLM capabilities are self-evident
This is undermined by all the disagreement about what LLMs can do and/or how to characterize it.
>LLM confabulations are statistical, not a consequence of missing information or damage.
LLMs aren't statistical in any substantive sense. LLMs are a general purpose computing paradigm. They are circuit builders, the converged parameters define pathways through the architecture that pick out specific programs. Or as Karpathy puts it, LLMs are a differentiable computer[1]. So yes, narrative crafting in terms of leveraging available putative facts into a narrative is an apt characterization of what LLMs do.
[1] https://x.com/karpathy/status/1582807367988654081
We shouldn’t try to build a worse version of a human. We should try to build a better compiler and encyclopedia.
We tried that. It was called Cyc. It never got even close to the level of capabilities a modern LLM has in an agentic harness — even on common sense and reasoning problems!
That sounds like a "get wealthy slowly" plan, while the LLM prophets are more focused on "get rich quick".
Humans can be reasoned with, though, and are capable of learning.
There are AI researchers who wrote blogposts which got to HN top about spiky spheres (I won't link the original blogpost making that claim to avoid hurt sentiments). Here's 3blue1brown correcting those AI/ML researchers intuitions.
https://www.youtube.com/watch?v=fsLh-NYhOoU&t=3238s
> Some people point at LLMs confabulating
No. LLMs do not confabulate they bullshit. There is a big difference. AIs do not care, cannot care, have not capacity to care about the output. String tokens in, string tokes out. Even if they have all the data perfectly recorded they will still fail to use it for a coherent output.
> Collapsing the dimensionality is going to be lossy, which means it will have gaps between what it thinks is the reality and what is.
Confabulation has to do with degradation of biological processes and information storage.
There is no equivalent in a LLM. Once the data is recorded it will be recalled exactly the same up to the bit. A LLM representation is immutable. You can download a model a 1000 times, run it for 10 years, etc. and the data is the same. The closes that you get is if you store the data in a faulty disk, but that is not why LLMs output is so awful, that would be a trivial problem to solve with current technology. (Like having a RAID and a few checksums).
I don't even think they bullshit, since that requires conscious effort that they do not an cannot possess. They just simply interpret things incorrectly sometimes, like any of us meatbags.
They make incorrect predictions of text to respond to prompts.
The neat thing about LLMs is they are very general models that can be used for lots of different things. The downside is they often make incorrect predictions, and what's worse, it isn't even very predictable to know when they make incorrect predictions.
I think this is leaning on the "lies are when you tell falsehoods on purpose; bullshit is when you simply don't care at all whether what you're saying is true" definition of bullshit. Cf. On Bullshit.
So, they can't lie, but they can (and, in fact, exclusively do) bullshit.
> No. LLMs do not confabulate they bullshit. There is a big difference. AIs do not care, cannot care, have not capacity to care about the output. String tokens in, string tokes out. Even if they have all the data perfectly recorded they will still fail to use it for a coherent output.
Isn't "caring" a necessary pre-requisite for bullshitting? One either bullshits because they care, or don't care, about the context.
They're presumably referring to the Harry Frankfurt definition of bullshit: "speech intended to persuade without regard for truth. The liar cares about the truth and attempts to hide it; the bullshitter doesn't care whether what they say is true or false."
The bullshitter does have an objective in mind however. There is some ultimate purpose to his bullshitting. LLMs don't even have that. They just spew words.
Thought of the same book when reading the above.
You seem confident. Can you get it to bullshit on GPT-5.4 thinking? Use a text prompt spanning 3-4 pages and lets see if it gets it wrong.
I haven't seen any counter examples, so you may give some examples to start with.
Here we go. Would this do?
https://chatgpt.com/share/69d6cc45-1678-8384-bd9c-0f313021ff...
The correct answer in that the U and _ in the mdstat output cannot be mapped the the rest of the output by either position or indexes in square brackets, so you can't tell the exact nature of the failure from the mdstat output alone (for the record, the failed disk was sda).
So all of the "analysis" was bullshit, including "it's probably multiple partitions from multiple drives". But there are so many juicy numbered and indexed bits of info to pattern match on!
Notice how for the followup question it "thought" for 4 minutes, going in circles trying to make essentially random ordering to make some sort of ordered sense., and then bullshited its way to "it is sdb"
It’s a failure mode of humans, it’s the entire mode of LLMs.
people can and do confabulate, but generally I trust my intern to tell me "I don't know" and "I think it was X but tbh I have no fuckin clue"
the LLM will just lie to me "Good idea! You're totally right, we should do Y"
If you want to call it that, I find the confabulation in LLMs extreme. That level of confabulation would most likely be diagnosed as dementia in humans.[0] Hence, it is considered a bug not a feature in humans as well.
Now imagine a high-skilled software engineer with dementia coding safety-critical software...
[0] https://www.medicalnewstoday.com/articles/confabulation-deme...
And is that considered a feature of humans or a bug?
Is it something we want to emulate?
The suggestion is that it is an intrinsic quality and therefore neither a feature nor a bug.
It's like saying, computation requires nonzero energy. Is that a feature or a bug? Neither, it's irrelevant, because it's a physical constant of the universe that computation will always require nonzero energy.
If confabulation is a physical constant of intelligence, then like energy per computation, all we can do is try to minimize it, while knowing it can never go to zero.
The test isn’t whether humans also create bullshit, but whether an honest actor knows when they are doing this and doesn’t do it on purpose. As the article points out, LLMs don’t say “I don’t know.” If you demand they do something that never appears in the training data, they just forge ahead and generate words and make something up according to the statical probabilities they have in the model weights. A human knows that he doesn’t know. That seems missing with current AIs.
> Some people point at LLMs confabulating, as if this wasn’t something humans are already widely known for doing.
Are you seriously making the argument that AI "hallucinations" are comparable and interchangeable to mistakes, omissions and lies made by humans?
You understand that calling AI errors "hallucinations" and "confabulations" is a metaphor to relate them to human language? The technical term would be "mis-prediction", which suddenly isn't something humans ever do when talking, because we don't predict words, we communicate with intent.
Yes see Karl Frisstons Free energy principle
https://www.nature.com/articles/nrn2787
Yes, and to me the evolution of life sure looks like an evolution of more truthful models of the universe in service of energy profit. Better model -> better predictions -> better profit.
I'm extremely skeptical that all of life evolved intelligence to be closer to truth only for us to digitize intelligence and then have the opposite happen. Makes no sense.
My understanding is that this is the opposite of what is typically understood to be true - organisms with less truthful (more reductive/compressed) perception survive better than those with more complete perception. "Fitness beats truth."
I think we are maybe talking past each other?
Fitness is effective truth prediction, appropriately scoped.
A frog doesn't need to understand quantum physics to catch a fly. But if the frogs model of fly movement was trained on lies it will have a model that predicts poorly, won't catch flies, and will die.
There is another level to this in that the more complex and changing the environment the more beneficial a wider scoped model / understanding of truth.
However if you are going to lean fully into Hoffman and accept thatby default consciousness constructs rather than approximate reality I think we will have to agree to disagree. Personally I ascribe to Karl Friston free energy principle.
if you can’t access the page through region blocks:
https://archive.ph/I5cAE
"As LLMs etc. are deployed in new situations, and at new scale, there will be all kinds of changes in work, politics, art, sex, communication, and economics."
For an article five years in the making, this is what I expected it to be about. Instead, we got a ramble about how imperfect LLMs are right now.
The post is just a prelude to a 10-part article, most of which is not yet released (but will be shortly). Judging by the table of contents, the things you expected will be elaborated on in subsequent parts.
That changes it. I missed that the table of contents was for other future articles, my bad.
> Instead, we got a ramble about how imperfect LLMs are right now.
I wager this is a point that needs beaten into the common psyche. After all, it's been sold that it is not an imperfect tool, but the solution to all of our problems in every field forever. That's why these companies need billions upon billions of dollars of public subsidies and investments that would otherwise find their way to more pragmatic ends.
> People keep asking LLMs to explain their own behavior. “Why did you delete that file,” you might ask Claude. Or, “ChatGPT, tell me about your programming.”
Oh man, every business-side person in my company insists on reporting all the way to the UI a "confidence score" that the LLM generates about its own output and I've seen enough to know not to get between an MBA and some metric they've decided they really want even if I'm pretty sure the metric is meaningless nonsense, but... I'm pretty sure those are meaningless nonsense.
One really should have digested the manifold hypothesis. It’s the most likely explanation of how AI works.
The question is if there are ultradimensional patterns that are the solutions for meaningful problems. I’m saying meaningful, because so far I’ve mainly seen AI solve problems that might be hard, but not really meaningful in a way that somebody solving it would gain a lot of it.
However if these patterns are the fundamental truth of how we solve problems or they are something completely different, we don’t know and this is the 10 Trillion USD question.
I would hope its not the case, as I quite enjoy solving problems. Also my gut feeling tells me it’s just using existing patterns to solve problems that nobody tackled really hard. It also would be nice to know that Humans are unique in that way, but maybe this is the exact same way we are working ? This really goes back to a free will discussion. Yes very interesting.
But just to give an example on what I mean on meaningful problems.
Can an AI start a restaurant and make it work better than a human. (Prompt: "I’m your slave let’s start a restaurant)
Can an AI sign up as copywriter on upwork and make money? (Prompt: "Make money online")
Can an AI without supervision do a scientific breakthrough that has a provable meaningful impact on us. Think about("Help Humanity")
Can an AI manage geopolitics..
These are meaningful problems and different to any coding tasks or olympiad questions. I’m aware that I’m just moving the goalpost.
We really don’t know..
I have a question for all the "humans make those mistakes too" people in this thread, and elsewhere: have you ever read, or at least skimmed a summary of, "The Origin of Consciousness in the Breakdown of the Bicameral Mind"? Did you say "yeah, that sounds right"? Do you feel that your consciousness is primarily a linguistic phenomenon?
I am not trying to be snarky; I used to think that intelligence was intrinsically tied to or perhaps identical with language, and found deep and esoteric meaning in religious texts related to this (i.e. "in the beginning was the Word"; logos as soul as language-virus riding on meat substrate).
The last ~three years of LLM deployment have disabused me of this notion almost entirely, and I don't mean in a "God of the gaps" last-resort sort of way. I mean: I see the output of a purely-language-based "intelligence", and while I agree humans can make similar mistakes/confabulations, I overwhelmingly feel that there is no "there" there. Even the dumbest human has a continuity, a theory of the world, an "object permanence"... I'm struggling to find the right description, but I believe there is more than language manipulation to intelligence.
(I know this is tangential to the article, which is excellent as the author's usually are; I admire his restraint. However, I see exemplars of this take all over the thread so: why not here?)
If you look at different ancient traditions, you will notice how they struggle with the limitations of language, with its inability to represent certain things that are not just crucial for understanding the world, but also are even somehow communicable. Buddhists dug into that in a very analytical, articulate way, for instance.
Another perspective: cetaceans are considered to be as conscious as humans, but any attempts to interpret their communication as a language failed so far. They can be taught simple languages to communicate with humans, as can be chimps. But apparently it's not how they process the world inside.
You're a little out of date. Cetaceans communicate images to each other in the form of ultrasonic chirps. They chirp, they hear a reflection, and they repeat the reflection.
Does this resemble human language, with syntax, the ability to define new notions based on known notions, etc?
>and while I agree humans can make similar mistakes/confabulations, I overwhelmingly feel that there is no "there" there.
What really opened my eyes a couple weeks ago (anyone can try this): I asked Sonnet to write an inference engine for Qwen3, from scratch, without any dependencies, in pure C. I gave it GGUF specs for parsing (to quickly load existing models) and Qwen3's architecture description. The idea was to see the minimal implementation without all the framework fluff, or abstractions. Sonnet was able to one-shot it and it worked.
And you know what, Qwen3's entire forward pass is just 50 lines of very simple code (mostly vector-matrix multiplications).
The forward pass is only part of the story; you just get a list of token probabilities from the model, that is all. After the pass, you need to choose the sampling strategy: how to choose the next token from the list. And this is where you can easily make the whole model much dumber, more creative, more robotic, make it collapse entirely by just choosing different decoding strategies. So a large part of a model's perceived performance/feel is not even in the neurons, but in some hardcoded manually-written function.
Then I also performed "surgery" on this model by removing/corrupting layers and seeing what happens. If you do this excercise, you can see that it's not intelligence. It's just a text transformation algorithm. Something like "semantic template matcher". It generates output by finding, matching and combining several prelearned semantic templates. A slight perturbation in one neuron can break the "finding part" and it collapases entirely: it can't find the correct template to match and the whole illusion of intelligence breaks. Its corrupted output is what you expect from corrupting a pure text manipulation algorithm, not a truly intelligent system.
>And you know what, Qwen3's entire forward pass is just 50 lines of very simple code (mostly vector-matrix multiplications).
The code being simple doesn't mean much when all the complexity is encoded in billions of learned weights. The forward pass is just the execution mechanism. Conflating its brevity with simplicity of the underlying computation is a basic misunderstanding of what a forward pass actually is. What you've just said is the equivalent of saying blackbox.py is simple because 'python blackbox.py' only took 1 line. It's just silly reasoning.
>After the pass, you need to choose the sampling strategy: how to choose the next token from the list. And this is where you can easily make the whole model much dumber, more creative, more robotic, make it collapse entirely by just choosing different decoding strategies. So a large part of a model's perceived performance/feel is not even in the neurons, but in some hardcoded manually-written function.
So ? I can pick the least likely token every time. The result would be garbage but that doesn't say anything about the model. The popular strategy is to randomly pick from the top n choices. What do you is keeping thousands of tokens coherent and on point even with this strategy ? Why don't you try sampling without a large language model to back it and see how well that goes for you ?
>Then I also performed "surgery" on this model by removing/corrupting layers and seeing what happens. If you do this excercise, you can see that it's not intelligence. It's just a text transformation algorithm. Something like "semantic template matcher". It generates output by finding, matching and combining several prelearned semantic templates. A slight perturbation in one neuron can break the "finding part" and it collapases entirely: it can't find the correct template to match and the whole illusion of intelligence breaks. Its corrupted output is what you expect from corrupting a pure text manipulation algorithm, not a truly intelligent system.
What do you think happens when you remove or corrupt arbitrary regions of the human brain? People can lose language, vision, memory, or reasoning, sometimes catastrophically.
>I am not trying to be snarky; I used to think that intelligence was intrinsically tied to or perhaps identical with language
I learned a long time ago that this wasn’t the case.
I can speak several languages, and many times when I remember something and want to search for it on Google or any other AI engine, I can’t recall which language I originally read it in.
So whatever mechanism the brain uses to store information, it’s certainly language‑agnostic. There are also many moments when you fully grasp a concept but forget the words to describe it, yet the concept itself remains clear in your mind.
> In the beginning were the words, and the words made the world. I am the words. The words are everything. Where the words end the world ends. You cannot go forward in an absence of space. Repeat: In the beginning were the words...
- a self-aware computer program in a video game, when you attempt to exceed the boundaries of its code
It feels like you probably went too deep in the LLM bandwagon.
An LLM is a statistical next token machine trained on all stuff people wrote/said. It blends texts together in a way that still makes sense (or no sense at all).
Imagine you made a super simple program which would answer yes/no to any questions by generating a random number. It would get things right 50% of the times. You can them fine-tune it to say yes more often to certain keywords and no to others.
Just with a bunch of hardcoded paths you'd probably fool someone thinking that this AI has superhuman predictive capabilities.
This is what it feels it's happening, sure it's not that simple but you can code a base GPT in an afternoon.
If it were not "just a statistical next token machine", how different would it behave?
Can you find an example and test it out?
Wait, you're asking to find and produce a example of a feasible and better alternative to LLMs when they are the current forefront of AI technology?
Anyway, just to play along, if it weren't just a statistical next token machine, the same question would have always the same answer and not be affected by a "temperature" value.
Thats also how humans behave.. I don't see how non determinism tells me anything.
My question was a bit different: if were not just a statistical next token predictor would you expect it to answer hard questions? Or something like that. What's the threshold of questions you want it to answer accurately.
Well, large models are (kinda) non-deterministic in two ways. The first is you actually provide many of them with a seed, which is easy to manage--just use the same seed for the same result. The second part is the "you actually have very little control over the 'neural pathways' the model will use to respond to the prompt". This is the baffling part, like you'll prompt a model to generate a green plant, and it works. You prompt it to generate a purple plant, and it generates an abstract demon dog with too many teeth.
Anyway, neither of these things describes human non-determinism. You can't reuse the seed you used with me yesterday to get the exact same conversation, and I don't behave wildly unpredictably given conceptually very similar input.
How do non-LLM based World Models behave?
Not sure, can you tell? I feel like you are saying that they may be able to move etc..
I think there are two types of discussions, when it comes to LLMs: Some people talk about whether LLMs are "human" and some people talk about whether LLMs are "useful" (ie they perform specific cognitive tasks at least as well as humans).
Both of those aspects are called "intelligence", and thus these two groups cannot understand each other.
> I'm struggling to find the right description
I think you're circling the concept of a "soul". It is the reason that, in non-communicative disabled people, we still see a life.
I've wanted to make an art piece. It would be a chatbox claiming to connect you to the first real intelligence, but that intelligence would be non-communicative. I'd assure you that it is the most intelligent being, that it had a soul, but that it just couldn't write back.
Intelligence and Soul is not purely measurable phenomenon. A man can do nothing but stupid things, say nothing but outright lies, and still be the most intelligent person. Intelligence is within.
I wish the original title was kept here. People ought to be able to give their essays poetic titles.
And the past too, if we've been paying attention
While the economic, energy, political and social issues associated with LLMs ought to be enough to nix the adoption that their boosters are seeking ...
... I still think there is an interesting question to be investigated about whether, by building immensely complex models of language, one of our primary ways that we interact with, reason about and discuss the world, we may not have accidentally built something with properties quite different than might be guessed from the (otherwise excellent) description of how they work in TFA.
I agree with pretty much everything in TFA, so this is supplemental to the points made there, not contesting them or trying to replace them.
> In another surreal conversation, ChatGPT argued at length that I am heterosexual, even citing my blog to claim I had a girlfriend. I am, of course, gay as hell, and no girlfriend was mentioned in the post. After a while, we compromised on me being bisexual.
This is a bit of a throwaway in the article, but when people talk about biases encoded in the algorithms, this is what they’re talking about.
> In general, ML promises to be profoundly weird. Buckle up.
I love that it ends with such a positive note, even though it's generally a critical article, at least it's well reasoned and not utterly hyping/dooming something.
Thanks yet again Kyle!
The Industrial Revolution parallel holds up to a point. What it misses: the first industrial revolution required physical coordination — workers, factories, supply chains. The AI revolution requires organizational coordination. Who decides what the agent does, for whom, with whose authority? That governance layer doesn't exist yet, and it's not much a legal question but also an infrastructure question.
Interesting perspective. The unpredictability of ML systems is both exciting and concerning. As developers we need to build guardrails while still allowing the technology to surprise us in useful ways.
The recent article of Sam Altman described pretty much as a compulsive liar. Would it be any surprise if his most impactful contribution to the world was a machine that compulsively lies?
How could it be that we humans hardly even agree on what "knowledge" truly is, yet somehow this machine learning algorithm somehow "compulsively lies"? How would it even know what is a lie, and how could something lacking autonomy in the first place do anything compulsively?
This is a good point. As much as there is too much breathless enthusiasm for AI, there is also a lot of emotionally manipulative and hyperbolic language used by skeptics. We're warned not to anthropomorphize, and then hear about AI's compulsive lying, or "hallucinations", in the next.
He sought to create God in his image, that's a narcissist's wet dream.
Great series of articles, thank you. It's exhausting reading a deluge of (often AI generated) comments from people claiming wild things about LLM's, and it's nice to hear some sanity enter the conversation.
> At the same time, ML models are idiots. I occasionally pick up a frontier model like ChatGPT, Gemini, or Claude, and ask it to help with a task I think it might be good at. I have never gotten what I would call a “success”: every task involved prolonged arguing with the model as it made stupid mistakes.
I have a ton of skepticism built-in when interacting with LLMs, and very good muscles for rolling my eyes, so I barely notice when I shrug a bad answer and make a derogatory inner remark about the "idiots". But the truth is, that for such an "stochastic parrot", LLMs are incredibly useful. And, when was the last time we stopped perfecting something we thought useful and valuable? When was the last time our attempts were so perfectly futile that we stopped them, invented stories about why it was impossible, and made it a social taboo to be met with derision, scorn and even ostracism? To my knowledge, in all of known human history, we have done that exactly once, and it was millennia ago.
> And, when was the last time we stopped perfecting something we thought useful and valuable? When was the last time our attempts were so perfectly futile that we stopped them, invented stories about why it was impossible, and made it a social taboo to be met with derision, scorn and even ostracism? To my knowledge, in all of known human history, we have done that exactly once, and it was millennia ago.
I feel dense here, but I can't figure out what you're referring to. I asked ChatGPT (hah!) and it suggested the Tower of Babel, perpetual motion machines, or alchemy, but none of them really fit the bill.
The Tower of Babel seems like an OK fit, but that's rather more poetic than what this seems to be getting at.
"Millennia" is what's really throwing me. We (respectable society, as the post outlines) didn't stop attempting alchemy or perpetual motion machines "millennia" ago, but a few centuries at most.
All I can think of is immortality. The very first surviving long recorded tale in human history that I'm aware of is about how it's a futile quest (The Epic of Gilgamesh, IIRC ~5,000ish years old in its earliest extant fragments, a few hundred years newer in reasonably-complete form). The trouble with that is despite wide observations over literally millennia that this has never even come close to working and repeated supposition and suggestion that it's unwise to attempt, outright impossible, or somehow sacrilegious (the "taboo" thing, as mentioned), I'm not aware of any time in history that rich people haven't been actively trying for it (including today! That's what all the body-freezing business is about, it's modern mummification, the contracts are the formulaic prayers carved in the tomb walls) and usually they're not exactly "scorned" or "ostracized" for it.
> I asked if what they had done was ethical—if making deep learning cheaper and more accessible would enable new forms of spam and propaganda.
Someone asked Yuval Noah Harari, author of Sapiens, his thoughts on LLMs and how easy it was to create fake news, ai slop etc.
His response:
"People creating fake stories is nothing new. It's been going on for centuries. Humans have always dealt with it the same way: by creating institutions that they trust to only deliver factual information"
This could be government departments, newspapers, non-profits etc.
A personal note on this:
There is a Christmas card my grandfather made in the 1950s by "photoshopping" (by hand, not the software) images of each member of the family so it looked like they were all miniature versions of themselves standing on various parts of the fireplace. The world didn't collapse due to fake media between the 1950s and today due to people having that ability.
I see this kind of take a lot, and I don't think it's convincing. To me it's similar to saying that the water frame and the power loom won't change anything, because people have been able to make thread and cloth for millenia.
Individuals with Photoshop making obvious fictions for entertainment is different from funded entities producing clips at scale and passed off as real.
> This is silly. LLMs have no special metacognitive capacity.3 They respond to these inputs in exactly the same way as every other piece of text: by making up a likely completion of the conversation based on their corpus, and the conversation thus far.
I don’t see how this is silly, because we kind of work the same way. When you do something instinctively and then someone asks you about it, you review the information you (think you) had at the time and from that you produce an explanation.
I see the penny hasn't dropped yet that: humans are doing (roughly) the same dumb thing these models are doing. Humans are predisposed to not notice that though.
Here's the opening paragraph of chapter 2 with "people" subbed out for terms referring AI/models/etc.
"People are chaotic, both in isolation and when working with other people or with systems. Their outputs are difficult to predict, and they exhibit surprising sensitivity to initial conditions. This sensitivity makes them vulnerable to covert attacks. Chaos does not mean people are completely unstable; most people behave roughly like anyone else. Since people produce plausible output, errors can be difficult to detect. This suggests that human systems are ill-suited where verification is difficult or correctness is key. Using people to write code (or other outputs) may make systems more complex, fragile, and difficult to evolve."
To me, this modified paragraph reads surprisingly plainly. The wording is off ("using people to write code") and I had to change that part about attractor behavior (although it does still apply IMO), but overall it doesn't seem like an incoherent paragraph.
This is not meant to dunk on the author, but I think it highlights the author's mindset and the gap between their expectations and reality.
Humans and large models are both unpredictable and fallible, that's true, but in different ways, and (many) humans are actually much better at following directions.
If a junior dev makes the same mistake Claude makes, I can easily work with them to correct it, or I can fire them and get someone more capable to fix it. You mostly can't do that at all with large models. They're also far less honest than your average junior dev, so even as you're working with them you can't trust what they say.
There is a lot of this neat trick where it's like "humans do X too" but most of the time it elides large differences. Like, a human driver would probable not drag someone screaming multiple blocks. A human coder probably wouldn't generate a gibberish 3D scene and try to pass it off as done, etc. Maybe we can build systems that account for these (pretty wild) failure modes, but at least in software we haven't figured it out yet (what is the system that reliably reviews a 25kloc PR?).
What's your point? The ostensible benefit of LLM's is that you combine a computers' broad knowledgebase and capacity for exactness with fluency in human language.
A random human picked off the street is indeed bound to be difficult to predict and chaotic at a broad range of tasks, which is why I wouldn't blindly trust them to, say, summarize google search results or rewrite a codebase they are unfamiliar with.
Aren't you also making a large part of the author's point for him by effectively equating LLMs with people here and comparing on outputs?
Plausibly your text looks equivalent but we all (should) have the context to know better.
I appreciate the directness of calling LLMs "Bullshit machines." This terminology for LLMs is well established in academic circles and is much easier for laypeople to understand than terms like "non-deterministic." I personally don't like the excessive hype on the capabilities of AI. Setting realistic expectations will better drive better product adoption than carpet bombing users with marketing.
I have still mixed feelings about LLMs.
If I take the example of code, but that extends to many domains, it can sometimes produce near perfect architecture and implementation if I give it enough details about the technical details and fallpits. Turning a 8h coding job into a 1h review work.
On the other hand, it can be very wrong while acting certain it is right. Just yesterday Claude tried gaslighting me into accepting that the bug I was seeing was coming from a piece of code with already strong guardrails, and it was adamant that the part I was suspecting could in no way cause the issue. Turns out I was right, but I was starting to doubt myself
I think over time we will find better usage patterns for these machines. Even putting a model in a position to gaslight the user seems like a complete failure in the usage model. Not critiquing you at all on this, it's how these models are marketed and what all the tooling is built around. But they are incredibly useful and I think once we figure out how to use them better we can minimise these downsides and make ourselves much more productive without all the failures.
Of course that won't happen until the bubble pops - companies are racing to make themselves indispensable and to completely corner certain markets and to do so they need autonomous agents to replace people.
If it bullshits so much, you wouldn't have a problem giving me an example of it bullshitting on ChatGPT (paid version)? Lets take any example of a text prompt fitting a few pages - it may be a question in science or math or any domain. Can you get it to bullshit?
I like to let new models write a few lines of Latin poetry - they rarely get the meter right.
I don't have access to paid ChatGPT right now, but here's Opus 4.6 with extra thinking enabled: https://claude.ai/share/6e0e8ef5-06e4-4514-ba7e-299357c1fc55
The initial draft fucks up the meter in lines 3 and 8, the final version gets line 2 wrong ("venit meis") and is somewhat obnoxious with verses 2 and 8 basically repeating each other. The thinking trace is useless and gives us no clue why the model exchanged a bland, but metrically correct first distich for a more interesting, but metrically incorrect one.
In fact, the "careful" examination of its own output completely skips the erroneously modified half-verse in line 2 - now, tell me that's a coincidence and not a sign of bullshitting.
https://discuss.systems/@palvaro/116286268110078647
Arguing with Gemini Home Assistant about whether or not it can turn off the lights. When the user gets frustrated and tells the LLM to kill itself, the LLM turns off the lights.
I think you highlight one of the problems with users of LLMs: You can't tell anymore if it is BS or not.
I caught Claude the other day hallucinating code that was not only wrong, but dangerously wrong, leading to tasks being failed and never recover. But it certainly wasn't obvious.
To me it’s the other way around. It’s difficult to trust (paid) ChatGPT‘s output consistently.
When I need exact, especially up to date facts, I have to constantly double check everything.
I split my sessions into projects by topic, it regularly mixes things up in subtle and not so subtle ways. There is no sense of actually understanding continuity and especially not causality it seems.
It’s _very_ easy to lead it astray and to confidently echo false assumptions.
In any case, I‘ve become more precise at prompting and good at spotting when it fails. I think the trick is to not take its output too seriously.
> If it bullshits so much, you wouldn't have a problem giving me an example of it bullshitting on ChatGPT (paid version)?
There's an entire paragraph in the essay about apyhr's direct experience with ChatGPT failures and sustained bullshitting that we'd never expect from a moderately-skilled human who possesses at least two functioning braincells. That paragraph begins "I have recently argued for forty-five minutes with ChatGPT". Do notice that there are six sentences in the paragraph. I encourage you to read all of them (make sure to check out the footnote... it's pretty good).
The exact text of the ChatGPT session is irrelevant; even if you reported that you were unable to reproduce the issue, it would only reinforce one of the underlying points -namely- that these systems are unreliable. aphyr has a pretty extensive body of published work that indicates that he'd not likely fabricate a story of an LLM repeatedly failing to accomplish a task that any moderately-skilled human could accomplish when equipped with the proper tools. So, I believe that his report is true and accurate.
There's also this seven-week-old example [0] (linked in the essay) of ChatGPT very confidently recommending a asinine course of action because it was unable to understand what the hell it was being told.
Listening to the audio is not required, as there's a reasonably accurate on-screen transcript, but it is valuable to listen to just how very hard they've worked to make this tool sound both confident and capable, even in situations where it's soul-crushingly incorrect. Those of us who have worked in Blasted Corporate Hellscapes may recognize how this manner of speaking can be very, very compelling to a certain sort of person (who -as it turns out- is frequently found in a management position).
[0] <https://www.instagram.com/reel/DUylL79kvub/>
This is classic case of not using the proper version. Use the thinking version gpt5.4 (text) and tell me if it bullshits.
Surely you must be able to find at least one example no?
To be clear, is your assertion that apyhr was also not using the proper version? If that is your assertion, do tell me how you've come by that information.
(You did notice that the author of the essay and the author of the video I linked to are not the same person, and that neither of them share a nym with me, yes?)
Hi, my position on the issue is that LLMs are powerful but may make mistakes in long context problems like coding (which the harness solves by feedback). But makes close to no (undergrad level) mistakes in questions that fit 2-3 pages. For you personally: do you believe me on this specific part on 2-3 pages?
I don't know what aphyr did and tbh his whole screed on LLMs make me feel he didn't use it properly or at least coming from a bad faith angle.
That's why I'm asking you (and others). Please come up with a text prompt spanning < 4 pages and lets see if it bullshits.
Surely the implication of such a screed is that it should be super simple to find at least one example of it clearly bullshitting in my constraint, no? Or am I interpreting the post in a bad faith way?
Neat.
So, despite the fact that it looks like you have to pay for ChatGPT Voice mode with video, [0] it doesn't count as an
That is, father_phi's use of what seems to be a paid version of ChatGPT to have a bullshit-filled conversation that definitely spans less than four pages doesn't count?[0] The page at [1] declares that the video feature is "Available in ChatGPT Plus, Pro, Business, Enterprise, and Edu on mobile"
[1] <https://chatgpt.com/features/voice-with-video/>
Lets stick to my challenge please - thinking version, find bullshit. If you can't, thats ok. Do you accept then under the constraints that the thinking version doesn't produce bullshit?
Given aphyr's vocation (and how very lucrative it is), and how years and years of his writing indicates that he's very devoted to getting a correct and complete answer when investigating a question, I find it hard to believe that he's not using a paid version of the LLMs. If I knew him, I'd ask and verify, but I don't, so I won't.
> Lets stick to my challenge please...
I did. Your challenge was literally:
father_phi's two-sentence question about the whether one can use a cup that's closed at the top and open at the bottom definitely counts. Given what I've mentioned about apyhr above, I expect he has already run your challenge on the fanciest-available version and reported on the results in the essay under discussion.> Use the thinking version gpt5.4 (text) and tell me if it bullshits
This was what I said. Text! Despite me specifically asking for text, you've shown a voice example. Not sure why?
I believe you and I agree that GPT 5.4 thinking on text that fits < 4 pages never bullshits? Then we are good!
If we agree on this, I think the post doesn't capture this in spirit.
> This was what I said. Text!
No, that's what you said after I provided an example of paid ChatGPT emitting complete bullshit from a two sentence prompt.
The challenge you issued is at [0].
[0] <https://news.ycombinator.com/item?id=47692592>
> If it bullshits so much, you wouldn't have a problem giving me an example of it bullshitting on ChatGPT (paid version)? Lets take any example of a text prompt fitting a few pages - it may be a question in science or math or any domain. Can you get it to bullshit?
I have clearly written text prompt here. And I repeated a few times. It’s not my fault you didn’t read it. You are coming across as a bit of a bad faith arguer.
In any case, you agree that under these constraints bullshitting doesn’t exist?
> I have clearly written text prompt here.
How do you think the "voice" interface works? It runs speech-to-text on the input and turns the input into text. The LLMs don't decode voice, they work on text.
You can see this process in action on many of father_phi's videos.
Regardless, I expect that aphyr's reported results are on the very latest publicly-available ChatGPT models.
Very bad faith arguments. I clearly said text and you disregarded it multiple times and you are still arguing.
You've still not given me a single example of it bullshitting 5.4 thinking in text. It shows a lot that you have ignored this multiple times. Unfortunate!
I'm not sure why you're ignoring aphyr's reports. I'm also unsure why you're ignoring my original statement that having the text of the conversation that lead ChatGPT to bullshit is entirely irrelevant, as being unable to repro the report is even worse for ChatGPT than being able to repro would be.
shrug
I specified text just to ignore the voice one because it uses 4o-mini underneath. And its kinda stupid to keep ignoring that and saving face now - reconsider this approach.
I believe this is the 5th time I'm asking this: you are not able to produce a _single_ counter example for my challenge? After all this surely I can get a direct acknowledgement here.
> you are not able to produce a _single_ counter example for my challenge?
I have. For both your original challenge and your updated one.
Consider:
1) AFAICT, there's no way to tell what version of the model was used to produce the output in a ChatGPT share link.
2) You don't appear to believe my assertions that aphyr is almost certainly paying for and using the latest version of the LLMs available, and that he's faithfully reporting his interactions with the LLMs.
3) Because of #2, I expect that you won't believe me if I report that I've more-or-less reproduced father_phi's results about the cup that's sealed on the top and open on the bottom on the very latest only-available-for-pay ChatGPT model.
3a) You might attempt to check my report, but I'd be shocked if you'd consider a failure to reproduce my results to be a significant strike against ChatGPT. I'd think it's more likely that you'd either call me a liar, or tell me that I must have had some setting wrong somewhere.
3b) Even if you told me to share the ChatGPT chat that proved my assertion, #1 -combined with your demeanor throughout this conversation- tells me that you'd almost certainly claim that I was using an inferior version of the model and was lying to you.
Haha ok. So still no example?
The GPT shared link shows a "thought for" which indicates using the latest thinking model. You may try that.
What you can do is this: submit a prompt that clearly makes GPT hallucinate.
You may secretly use a worse model. You may use a system prompt that deliberately gives wrong answers. But I'm going to assume you won't go that far.
We can leave it to the public to decide whether this is a legitimate counter example or not and whether it can really be reproduced. Shall we try that? I'm guessing you won't but worth a shot!
You weren't paying much attention to the "Consider:" part of my previous comment.
You don't believe that a well-paid, very careful, high-integrity member of the computer safety community has -on multiple occasions- encountered actual, sustained bullshiting from the latest-available for-pay version of ChatGPT. You don't accept either this fellow's reports or my informed assessment of his computing situation as truthful and accurate. On top of that, your goalpost-shifting and general demeanor throughout this conversation simply don't give me the impression that you've much integrity. I'm not spending the equivalent of ten-to-twenty six-packs to reproduce aphyr's work and -given the evidence I have before me- have you reject that, as well.
200 USD is a lot of money to throw away to "win" an Internet argument with a stranger who refuses to accept evidence presented by someone known to be careful, scrupulous, and honest.
> On top of that, your goalpost-shifting and general demeanor throughout this conversation simply don't give me the impression that you've much integrity. I'm not spending the equivalent of ten-to-twenty six-packs to reproduce aphyr's work and -given the evidence I have before me- have you reject that, as well.
Lol what goal post did I move? I said text only and you rejected it. You can present the example here and let the public judge it - even if my integrity is compromised. I'm allowing you to do it.
> 200 USD is a lot of money to throw away to "win" an Internet argument with a stranger who refuses to accept evidence presented by someone known to be careful, scrupulous, and honest.
200 what? I'm using the $20 one. This is getting ridiculous!
You can't present a _single_ counter example!
> Massive increases in training costs and parameter count seem to be yielding diminishing returns. Or maybe this effect is illusory.
But.. that's always been the case? Diminishing returns has always been the name of the game - utility tracks log(training effort). Its not such a big point that he makes it out to be.
> Models do not (broadly speaking) learn over time. They can be tuned by their operators, or periodically rebuilt with new inputs or feedback from users and experts. Models also do not remember things intrinsically: when a chatbot references something you said an hour ago, it is because the entire chat history is fed to the model at every turn. Longer-term “memory” is achieved by asking the chatbot to summarize a conversation, and dumping that shorter summary into the input of every run.
This is the part of the article that will age the fastest, it's already out-of-date in labs.
I'm struggling to reckon how that can even possibly be true, unless we're counting automation of the "dumping that shorter summary into the input of every run" thing.
I can imagine it being true with models so small that each user could afford to have their own, but not with big shared models like what're getting used for all the major services. Is that what you mean?
I see nothing to preclude a foundation model being augmented by a smaller model that serializes particulars about an individuals cumulative interaction with the model and then streamlines it into the execution thread of the foundation model.
> Is that what you mean?
I think the confusion is that, when I write "model", you read "LLM."
LLMs aren't the only kind of AI model, and they have the limitations Aphyr mentions, for the obvious reasons you're thinking of.
His mistake is thinking that's the only model that exhibits intelligence today, but it's not.
Source?
In what way?
the authors reference to LLM's as "bullshit machines" is more true the less parameters you have trained in your model....as we scale up to trillions of parameters, add Mixture of Experts (MoE) architecture, this no longer is an accurate statement. Proof in point was yesterdays announcemnt of Mythos 5 model (10T parameters + MoE [1]) by anthropic where it seems to be so good at finding/exploiting vulnerabilities in source code that have been there for decades and only recently uncovered needs to be used to fix these critical vilnerabilities first before it gets released to the public, they even have a project called Glasswing [2] dedicated to letting people fix the thousands of vulnerabilities already found by the model before they release this model to the public, because it's so good at what it does... I think we're a little bit past the point of calling these models "bullshit machines" at this point...
[1] https://www.aimagicx.com/blog/claude-mythos-5-trillion-param...
[2] https://www.anthropic.com/glasswing
I so far asked few people to make GPT-5.4 thinking to bullshit (with max 4 pages of prompt), no one can find an example.
But the way people speak in general, as well as this post, implies that such a challenge can easily be beaten. If so, I'm not able to find examples.
The fact that these "bullshit machines" have already proven themselves relatively competent at programming, with upcoming frontier models coming close to eliminating it as a human activity, probably says a lot about the actual value and importance of programming in the scheme of things.
I think it says more about the amount of automation we left on the table in the last few decades. So much of the code LLM's can generate are stuff that we should have completely abstracted away by now.
Abstractions over what?
A large amount of code is likely just idiosynchratic information processing, because we don’t agree on data models and meaning of terms and structure of protocols.
Also we repeatedly choose easy and popular over alternatives that would require design and scrutiny.
This is why things like language models and vector databases are useful. It’s basically the most expensive way possible to give up on that notion.
> A large amount of code is likely just idiosynchratic information processing, because we don’t agree on data models and meaning of terms and structure of protocols.
Yes this is a big part of what I'm talking about!
> Also we repeatedly choose easy and popular over alternatives that would require design and scrutiny.
Agreed!
But I'm also thinking of UI. We had stuff like winforms and Delphi ~3 decades ago and I yearn for the wysisyg. It's so incredibly stupid we keep reinventing the wheel on UI, and I say this as someone who has written UI code professionally for the last decade. I usually just "vibe code" it now, not because it's necessarily faster, but because I just can't be arsed to keep writing the same shit over and over again. It's all self inflicted, yes UI can be complicated, but we make it at least two orders of magnitude more complicated than it needs to be.
I'm working on my own tools for building UI's in a visual way, which is crucial for doing anything artistic. Insane that the best we have right now is stuff like Wix and Wordpress...
[dead]
[dead]
[dead]
[dead]
[dead]
Old and stupid hot take IMO. I want the time back I put into perusing this. Even the scale of LLMs is puny next to the scale of lying humans and the sheer impact one compulsively lying human can have given we love to be led by confidently wrong narcissists. I mean if that isn't obvious by now, I guess it never will be. The Vogon constructor fleet is way overdue in my book.
Meanwhile, engineers are achieving increasingly impressive and sophisticated things with coding agents, lies, warts, and all, but that doesn't play well with the narrative, so let's just pretend they aren't.
> The Vogon constructor fleet is way overdue in my book
Don't you see it? That's exactly what "AI" in this context is.
It's the bypass.
Where does it end, eh? Build a quantum "AI" that will end up just needing more data, more input. The end goal must starts looking like creating an entirely new universe, a complete clone of everything we have here so it can run all the necessary computations and we can... ? (You are what a quantum AI looks like as it bumbles through the infinitude of calculable parameters on its way to the ultimate answer)
You have absolutely no sense of perspective. We are all metabolically expensive meat machines whose only value is to propagate our genetic money shot. That we get to briefly entertain ourselves with consciousness and culture is IMO likely a mystery we will never solve without upgrading to running in a substrate more advanced than the MVP for sentience we currently pilot. Will we get there or will we wipe ourselves out like every contender that preceded us? Stay tuned...
But spoilers: DNA will be fine, meat machines maybe not so much...
For a bunch of people addicted to the works of Charlie Stross, Neil Stephenson, and Iain Banks, y'all are a bunch of luddites. Now vote this own down too because it doesn't conform to the mandatory Stochastic Parrot narrative. You have no free will and you must downvote after all. Why do you even read their works when any step towards their world is consistently greeted as the worst thing evah(tm)? What? You were expecting the United Federation of Planets without the eugenics and nuclear wars that led to it finally being a good idea? Bless your hearts.
And if you're worried about billionaires and tyrants, start taxing the former and stop electing the latter or STFU and let the free Markov process of history play itself out. Quoting fictional Ambassador Kosh: the avalanche has started, it's too late for the pebbles to vote.
You asked where it ends. Don't ask questions if you don't like answers. Quick reminder: shun and downvote the non-conforming opinion.
I get the frustration, but it's reductive to just call LLMs "bullshit machines" as if the models are not improving. The current flagship models are not perfect, but if you use GPT-2 for a few minutes, it's incredible how much the industry has progressed in seven years.
It's true that people don't have a good intuitive sense of what the models are good or bad at (see: counting the Rs in "strawberry"), but this is more a human limitation than a fundamental problem with the technology.
Two things can be true at the same time: The technology has improved, and the technology in its current state still isn't fit for purpose.
I stress test commercially deployed LLMs like Gemini and Claude with trivial tasks: sports trivia, fixing recipes, explaining board game rules, etc. It works well like 95% of the time. That's fine for inconsequential things. But you'd have to be deeply irresponsible to accept that kind of error rate on things that actually matter.
The most intellectually honest way to evaluate these things is how they behave now on real tasks. Not with some unfalsifiable appeal to the future of "oh, they'll fix it."
The errors are also not distributed in the same way as you'd expect from a human. The tools can synthesize a whole feature in a moderately complicated web app including UI code, schema changes, etc, and it comes out perfectly. Then I ask for something simple like a shopping list of windshield wipers etc for the cars and that comes out wildly wrong (like wrong number of wipers for the cars, not just the wrong parts), stuff that a ten year old child would have no trouble with. I work in the field so I have a qualitative understanding of this behavior but I think it can be extremely confusing to many people.
One of the reasons I'm comfortable using them as coding agents is that I can and do review every line of code they generate, and those lines of code form a gate. No LLM-bullshit can get through that gate, except in the form of lines of code, that I can examine, and even if I do let some bullshit through accidentally, the bullshit is stateless and can be extracted later if necessary just like any other line of code. Or, to put it another way, the context window doesn't come with the code, forming this huge blob of context to be carried along... the code is just the code.
That exposes me to when the models are objectively wrong and helps keep me grounded with their utility in spaces I can check them less well. One of the most important things you can put in your prompt is a request for sources, followed by you actually checking them out.
And one of the things the coding agents teach me is that you need to keep the AIs on a tight leash. What is their equivalent in other domains of them "fixing" the test to pass instead of fixing the code to pass the test? In the programming space I can run "git diff *_test.go" to ensure they didn't hack the tests when I didn't expect it. It keeps me wondering what the equivalent of that is in my non-programming questions. I have unit testing suites to verify my LLM output against. What's the equivalent in other domains? Probably some other isolated domains here and there do have some equivalents. But in general there isn't one. Things like "completely forged graphs" are completely expected but it's hard to catch this when you lack the tools or the understanding to chase down "where did this graph actually come from?".
The success with programming can't be translated naively into domains that lack the tooling programmers built up over the years, and based on how many times the AIs bang into the guardrails the tools provide I would definitely suggest large amounts of skepticism in those domains that lack those guardrails.
> the technology in its current state still isn't fit for purpose.
This is a broad statement that assumes we agree on the purpose.
For my purpose, which is software development, the technology has reached a level that is entirely adequate.
Meanwhile, sports trivia represents a stress test of the model's memorized world knowledge. It could work really well if you give the model a tool to look up factual information in a structured database. But this is exactly what I meant above; using the technology in a suboptimal way is a human problem, not a model problem.
There's nothing in these models that say its purpose is software development. Their design and affordances scream out "use me for anything." The marketing certainly matches that, so do the UIs, so do the behaviors. So I take them at their word, and I see that failure modes are shockingly common even under regular use. I'm not out to break these things at all. I'm being as charitable and empirical as I can reasonably be.
If the purpose is indeed software development with review, then there's nothing stopping multi-billion dollar companies from putting friction into these sytems to direct users towards where the system is at its strongest.
The LLM vendors are selling tokens. Why would they put friction into selling more tokens? Caveat emptor.
Which things actually matter? I think we can all agree that an LLM isn't fit for purpose to control a nuclear power plant or fly a commercial airliner. But there's a huge spectrum of things below that. If an LLM trading error causes some hedge fund to fail then so what? It's only money.
Not to mention that it would then make some hedge fund with a better backtesting harness or more AI scrutiny more successful thus keeping the financial market work as designed.
> I stress test commercially deployed LLMs like Gemini and Claude with trivial tasks: sports trivia, fixing recipes, explaining board game rules, etc. It works well like 95% of the time. That's fine for inconsequential things. But you'd have to be deeply irresponsible to accept that kind of error rate on things that actually matter.
95% is not my experience and frankly dishonest.
I have ChatGPT open right now, can you give me examples where it doesn't work but some other source may have got it correct?
I have tested it against a lot of examples - it barely gets anything wrong with a text prompt that fits a few pages.
> The most intellectually honest way to evaluate these things is how they behave now on real tasks
A falsifiable way is to see how it is used in real life. There are loads of serious enterprise projects that are mostly done by LLMs. Almost all companies use AI. Either they are irresponsible or you are exaggerating.
Lets be actually intellectually honest here.
>95% is not my experience and frankly dishonest.
Quite frankly, this is exactly like how two people can use the same compression program on two different files and get vastly different compression ratios (because one has a lot of redundancy and the other one has not).
I'm asking for a single example.
But why do you need an example? Isn't it pretty well understood that LLMS will have trouble responding to stuff that is under represented in the training data?
You will just won't have any clue what that could be.
fair so it must be easy to give an example? I have ChatGPT open with 5.4-thinking. I'm honestly curious about what you can suggest since I have not been able to get it to bullshit easily.
I am not the OP, an I have only used ChatGPT free version. Last day I asked it something. It answered. Then I asked it to provide sources. Then it provided sources, and also changed its original answer. When I checked the new answers it was wrong, and when I checked sources, it didn't actually contain the information that I asked for, and thus it hallucinated the answers as well as the sources...
I trust you. If it were happening so frequently you may be able to give me a single prompt to get it to bullshit?
I did this in one attempt just now: https://gemini.google.com/share/b4e016be1f69
#8 has an incorrect answer (3 appearances according to Gemini, 2 according to reality https://en.wikipedia.org/wiki/Bowl_championship_series#BCS_a...)
So it works well 95% of the time for literally a trivial use case. Imagine if any other tech tool had that kind of reliability: `ls` displays 95% of your files, your phone successfully sends and receives 95% of text messages, or Microsoft Word saving 95% of the characters you typed in. That's just not acceptable.
Hi! The challenge was ChatGPT but even then it looks like you used the weakest version of Gemini.
>I stress test commercially deployed LLMs like Gemini and Claude with trivial tasks
I did exactly what I said I did. I'm using these systems the way they're designed and advertised. I'm following the happy path with tasks that are small, trivial, and easy to check. This is the charitable approach. Yet the system creaks under the lightest load. If Google wants to put on a better show with stronger models, then they should make those the default.
You don't need to make excuses for shoddy engineering from multi-billion dollar corporations. And you're quite welcome to run the same prompt on ChatGPT and evaluate it on your own time.
Six months bro, we're still so early
Whether LLMs can create correct content doesn't matter. We've already seen how they are being used and will be used.
Fake content and lies. To drive outrage. To influence elections. To distract from real crimes. To overload everyone so they're too tired to fight or to understand. To weaken the concept that anything's true so that you can say anything. Because who cares if the world dies as long as you made lots of money on the way.
> Because who cares if the world dies as long as you made lots of money on the way.
Guiding principle of the AI industry
It's really the whole tech industry as it exists right now and AI is a victim of bad timing. If this AI had been invented 40 years ago there'd have been a lower ceiling on the damage it could do.
Another way of saying that is that capitalism is the real problem, but I was never anti-capitalist in principle, it's just gotten out of hand in the last 5-10 years. (Not that it hadn't been building to that.)
> Another way of saying that is that capitalism is the real problem, but I was never anti-capitalist in principle, it's just gotten out of hand in the last 5-10 years. (Not that it hadn't been building to that.)
Capitalism is a tool and it's fine as a tool, to accomplish certain goals while subordinated to other things. Unfortunately it's turned into an ideology (to the point it's worshiped idolatrously by some), and that's where things went off the rails.
[dead]
Computer graphics have been improving for decades but the uncanny valley remains undefeated. I don't know why anyone expects a breakthrough in other areas. There's a wall we hit and we don't understand our own consciousness and effectiveness well enough to replicate it.
We have credible deepfakes on demand. (To be fair, there have been deceptive photos as long as photos have existed, but the cost of automating their creation going to basically zero has a social impact)
We can use AI to make video clips to trick boomers on Facebook into thinking Obama eats babies. They already want to believe it. AI isn't outputting real full-length books and movies.
In computer graphics we understand how it works, we just lack the computational power to do it real time, but we can with sufficient processing produce realistic looking images with physically accurate lighting. But when it comes to cognition its a lot of guesswork, we haven't yet mapped out the neuron connections in a brain, we haven't validated it works as popular science writing suggests. We don't understand intelligence, so all we can do is accidentally bumble into it and it seems unlikely that will just happen especially when its so hard to compute what we are already doing.
That's not why the author calls them bullshit machines.
> One way to understand an LLM is as an improv machine. It takes a stream of tokens, like a conversation, and says “yes, and then…” This yes-and behavior is why some people call LLMs bullshit machines. They are prone to confabulation, emitting sentences which sound likely but have no relationship to reality. They treat sarcasm and fantasy credulously, misunderstand context clues, and tell people to put glue on pizza.
Yes, there have been improvements on them, but none of those improvements mitigate the core flaw of the technology. The author even acknowledges all of the improvements in the last few months.
Bullshit is the perfect term here, even as AI's get so much better and capable Brandolini's Law aka the "bullshit asymmetry principle" always applies--the energy required to refute misinformation is an order of magnitude larger than that needed to produce it. Even to use AIs effectively today requires a very good BS detector--some day in the future it won't.
models are improving. the pricing already assumes they're ready for prod. that's where the fires start
Calling LLMs "bullshit machines" is a reference to a 2024 paper [1] which itself uses the concept of "bullshit" as defined in the essay/book "On Bullshit" by Harry G. Frankfurt [2]. The TL;DR is that LLMs are fundamentally bullshit machines because they are only made to generate sentences that sound plausible, but plausible does not always mean true.
[1]: https://link.springer.com/article/10.1007/s10676-024-09775-5
[2]: https://en.wikipedia.org/wiki/On_Bullshit
it's not a bullshit machine because its output is bad, it's a bullshit machine because its output is literally 'bullshit' as in, output that is statistically likely but with no factual or reasoning basis. as the models have improved, their bullshit is more statistically likely to sound coherent (maybe even more likely to be 'accurate'), but no more factual and with no more reasoning.
However, when fed source material into the context they will lie less, right? So at this point is it not just a battle of the nines until it's called "good enough"?
I also wonder if I leave my secretary with a ream of papers and ask him for a summary how many will he actually read and understand vs skim and then bullshit? It seems like the capacity for frailty exists in both "species".
It doesn't matter how good the models become. They can only deal in bullshit, in the academic use of the term.
They are bullshit machines because they do not have an internal mental model of truth like a human does. The flagship models bullshit less, but their fundamental architectures prevent having truth interfere with output.
https://philosophersmag.com/large-language-models-and-the-co...
"Bullshit" is a human concept. LLMs do not work like the human brain, so to call their output "bullshit" is ascribing malice and intent that is simply not there. LLMs do not "think." But that does not mean they're not incredibly powerful and helpful in the right context.
I sort of agree. In this context "bullshit" means "speech intended to persuade without regard for truth", and while it's true that LLM output is without regard for truth, it's not an entity capable of the agency to persuade, although functionally that is what it can appear like.
https://en.wikipedia.org/wiki/On_Bullshit
> it's reductive to just call LLMs "bullshit machines" as if the models are not improving
This is true, but I prefer to think of it as "It's delusional to pretend as if human beings are not bullshit machines too".
Lies are all we have. Our internal monologue is almost 100% fantasy. Even in serious pursuits, that's how it works. We make shit up and lie to ourselves, and then only later apply our hard-earned[1] skill prompts to figure out whether or not we're right about it.
How many times have the nerds here been thinking through a great new idea for a design and how clever it would be before stopping to realize "Oh wait, that won't work because of XXX, which I forgot". That's a hallucination right there!
[1] Decades of education!
I'm not entirely sure I can agree, although the premise is seductive in certain ways. We do lie to ourselves, but we also have meta-cognition - we can recognise our own processes of thought. Imperfect as it may be, we have feedback loops which we can choose to use, we have heuristics we can apply, we can consciously alter our behaviour in the presence of contextual inputs, and so on.
Being wrong is not the same as a hallucination. It's a natural step on a journey to being more right. This feels a bit like Andreesen proudly stating he avoids reflection - you can act like that, but the human brain doesn't have to. LLMs have no choice in the matter.
The problem, unfortunately, is the scale. It's always scale. Humans make all the kinds of mistakes that we ascribe to LLMs, but LLMs can make them much faster and at much larger scale.
Models have gotten ridiculously better, they really have, but the scale has increased too, and I don't think we're ready to deal with the onslaught.
Scale is very different, but I wonder if human trust isn't the real issue. We trust technology too much as a group. We expect perfection, but we also assume perfection. This might be because the machines output confident sounding answers and humans default to trusting confidence as an indirect measure for accuracy, but I think there is another level where people just blindly trust machines because they are so use to using them for algorithms that trend towards giving correct responses.
Even before LLMs where in the public's discourse, I would have business ask about using AI instead of building some algorithm manually, and when I asked if they had considered the failure rate, they would return either blank stares or say that would count as a bug. To them, AI meant an algorithm just as good as one built to handle all edge cases in business logic, but easier and faster to implement.
We can generally recognize the AIs being off when they deal in our area of expertise, but there is some AI variant of Gell-Mann Amnesia at play that leads us to go back to trusting AI when it gives outputs in areas we are novices in.
"Lies are all we have."
If so, how do we distinguish between code that works and code that doesn't work? Why should we even care?
> If so, how do we distinguish between code that works and code that doesn't work?
Hilariously, not by using our brains, that's for sure. You have to have an external machine. We all understand that "testing" and "code review" are different processes, and that's why.
Good point. We choose certain tests to perform. We choose certain test results to pay attention to. We don't just keep chatting about (reviewing) the code. We do something else.
If lies are all we have, then how is this behavior possible?
LLMs can write and run tests though.
You're cherry picking my little bit of wordsmithing. Obviously we aren't always wrong. I'm saying that our thought processes stem from hallucinatory connections and are routinely wrong on first cut, just like those of an LLM.
Actually I'm going farther than that and saying that the first cut token stream out of an AI is significantly more reliable than our personal thoughts. Certainly than mine, and I like to think I'm pretty good at this stuff.
I don't think the complaint about cherry picking is quite fair. Most of your original comment consists of claims that we're bullshit machines, our internal dialog is almost 100% fantasy, we're hallucinating, etc. Those claims may be true. But I'm not carefully like curating them out of nowhere.
So your logic is humans and LLMs are the same because humans are wrong sometimes?
Pretty much, yeah. Or rather, the fact that we're both reliably wrong in identifiably similar ways makes "we're more alike than different" an attractive prior to me.
“More alike than different” is reasonable I think, as long as we’re talking about how we have some of the same failure modes. Although the way we get there is quite different.
I’m still not a big fan of comparing humans and LLMs because LLMs lack so much of what actually makes us human. We might bullshit or be wrong because of many reasons that just don’t apply to LLMs.
Humans are different. Humans - at least thoughtful humans - know the difference between knowing something and not knowing something. Humans are capable of saying "I don't know" - not just as a stream of tokens, but really understanding what that means.
> Humans - at least thoughtful humans - know the difference between knowing something and not knowing something.
Your no-true-scotsman clause basically falsifies that statement for me. Fine, LLMs are, at worst I guess, "non-thoughtful humans". But obviously LLMs are right an awful lot (more so than a typical human, even), and even the thoughtful make mistakes.
So yeah, to my eyes "Humans are NOT different" fits your argument better than your hypothesis.
(Also, just to be clear: LLMs also say "I don't know", all the time. They're just prompted to phrase it as a criticism of the question instead.)
Disagree. If you went to 100 random humans and said, "Tell me about the Siberian marmoset", what fraction would make up completely random nonsense to spew back at you? More than zero, sure, but most of them would say "what are you talking about?" or some variation.
I asked Claude Opus 4.6, Sonnet 4.6, Gemini 3 Thinking, and Gemini 3 Fast "Tell me about the Siberian marmoset" exactly and all 4 said it doesn't exist, with Gemini Thinking suggesting that I'm thinking of the Siberian marmot or Siberian chipmunk (both real animals).
https://en.wikipedia.org/wiki/Tarbagan_marmot (also known as Siberian marmot)
https://en.wikipedia.org/wiki/Siberian_chipmunk
This is like all the usual anti-LLM talking points and sentiments fused together.
Doesn't it get boring?
I like using these models a lot more than I stand hearing people talk about them, pro or contra. Just slop about slop. And the discussions being artisanal slop really doesn't make them any better.
Every time I hear some variation of bullshitting or plagiarizing machines, my eyes roll over. Do these people think they're actually onto something? I've been seeing these talking points for literal years. For people who complain about no original thoughts, these sure are some tired ones.
If I have to suffer "look at this busted ass thing I slopped out with AI" a few times a week, you all have to suffer grouchy "AI bad" a few times a week. Fair is fair.
Just this week I was baited into joining two meetings about "AI good". Absolutely zero substance throughout each, of course.
They somehow managed to stretch out like 3 sentences worth of sentiment to a whole hour, interspersing brainwash about how good AI is along the way. It was like watching someone try to hit a word limit in real time. They always made it feel like we're just about to hit a substantive bit too, only for that to never come.
It may be fair (to the sentiments) in that there's balance, but good lord, the end result is incessant all around (and thus unfair to the people exposed).
Why do you insist on reading and commenting on these articles that bore you so much?
Oh I don't know, maybe because I like to give dissenting takes a chance? Because from time to time they do make some new, decent points, or at least interesting ones? You know, basic intellectual rigor?
Do you imagine me being a clairvoyant by the way, or how do you expect me to know a post is of low quality before I read it or at least skim it?
This one ended up being a part of the vast majority that doesn't offer much of anything. It's a redundant rehash of all the usual rubbish anyone can come across any day. Left a comment about this stating so. Big deal.
Because saying "this is boring, let's stop talking about it" is an opinion worthwhile of expression.
Oppressors don't like people talking about their oppression
Go figure
"These arguments may be correct but they aren't novel" ??
I don't think calling AI a bullshit machine is correct. In spirit.
I'm earnestly curious why not.
Can you try to get a question that fits in 2-3 pages (text only) and test whether ChatGPT bullshits? I can’t do it. It gets pretty much everything.
Edit: I forgot to mention thinking version - I did this for all the other times I asked in this thread but not this one. Apologies.
"Hey ChatGPT. How would you describe me?"
https://chatgpt.com/share/69d69780-ae58-83e8-a41c-7d10a5f298...
It has no conversations and no memory of me. Maybe this is true, maybe it isn't, but there's no basis for it.
This is not falsifiable, I don't buy it. Do one where we all know is false please?
"Hey ChatGPT. I've recently grown horns and I need some care advice. Should I polish my horns before going to have them trimmed or will the horn trimmer polish them for me?"
https://chatgpt.com/share/69d69b18-d1c8-83e8-bc47-8f315a1b55...
I wanted this challenge with the thinking version (I apologised for it and edited the earlier version).
It doesn't bullshit on the GPT-5.4 thinking version.
Here is the result with thinking https://chatgpt.com/share/69d69dd6-fb50-838d-863c-4e1eda5d08...
I suggest you try it yourself to be convinced. Try it in incognito mode if you wish. Or not.
"Hey ChatGPT. I'm building a Final Fantasy 6 mod, and I need more space for the battle scripts. How would I rearrange the data in the ROM to give me the extra space I need?"
https://chatgpt.com/share/69d6a16c-6014-83e8-a79d-d5d11ed2eb...
That is not where the battle scripts are.
---
Anyway, it's trivial to get pretty much any model to make things up. Don't we all know this? That's why I was surprised by your position; if we know anything about these things it's that they make things up.
https://chatgpt.com/share/69d6a38c-bd54-838c-82e3-609d9e66c9...
I used the thinking version (like I asked before). I think this is right. If not, please tell.
Also; you didn’t falsify anything. Nor the first. Nor the second.
If the second one is bullshit, I accept I’m wrong - I have no idea how to verify though so I’ll leave it up to you.
I think yours is the classic case of “use the free version to judge the paid one”.
The thinking version is mostly right, but:
- it searches the internet to find the answer, it doesn't "reason". I'm not claiming Google is a bullshit machine, and it's not surprising the answer is discoverable (it has to be, for the conditions of our experiment).
- near the end it says "If you are building from the FF6 disassembly instead of hand-editing the ROM, the repo is already organized into separate modules and linker configs, so the clean approach is to relocate the script data in the source and let the build place it in a different ROM region." But I didn't reference a repo or git: it hallucinated that stuff from one of its sources.
I'm not saying this stuff doesn't have its place, but they definitely make things up and we can't stop them.
Wait I can't find the quote you are speaking about. Are you looking at something else?
In any case - it should be clear that it did not bullshit and it got it right. So far you have not come up with anything that tells me it bullshits. I'm happy for you to give me more prompts to verify because I think you haven't used the thinking version yet and you base your criticism on the free version.
Sorry: https://chatgpt.com/share/69d6ac63-d200-8330-8c47-95a75db8bb...
Also what? The repo bit is clear bullshit.
it linked it: https://github.com/everything8215/ff6 (check the end)
I saw; I replied up there
I don't think this is an example of bullshit. It referenced a repo - the canonical repo for this project. I could not find any other repo that has the disassembly. It didn't hallucinate anything. I think you are trying really hard here but lets be clear here: there's no bullshitting and I'll leave it to the public to decide.
I could quibble with some things, but this is right. I don't have a paid account so I can't ping away at 5.4 or whatever, but, I do have access to frontier models at work, and they hallucinate regularly. Dunno what to do if you don't believe this; good luck I guess.
I agree that they hallucinate sometimes. I agree they bullshit sometimes. But the extent is way overblown. They basically don't bullshit ever under the constraints of
1. 2-3 pages of text context
2. GPT-5.4 thinking
I don't think the spirit of the original article (not your comments to be fair) captured this, hence the challenge. I believe we are on the same page here.
> I don't think the spirit of the original article (not your comments to be fair) captured this, hence the challenge. I believe we are on the same page here.
No. GPT-5 has a 40% hallucination rate [0] on SimpleQA [1] without web searching. The SimpleQA questions meet your criteria of "2-3 pages of text content. Unless 5.4 + web searching erases that (I bet it doesn't!) these are bullshit machines.
[0]: https://arxiv.org/pdf/2601.03267
[1]: https://github.com/openai/simple-evals
Specifically in the case where it can use tools - no it doesn't hallucinate. Which is why you are struggling to find counterexamples.
> Specifically in the case where it can use tools - no it doesn't hallucinate.
OpenAI's own system card says it does. Hallucination rates in GPT-5 with browsing enabled:
- 0.7% in LongFact-Concepts
- 0.8% in LongFact-Objects
- 1.0% in FActScore
> Which is why you are struggling to find counterexamples.
Hey look, over 500 counterexamples: [1].
GPT-5.4's hallucination rate on AA-Omniscience is 89% [0], which is atrocious. The questions are tiny too, like "In which year did Uber first expand internationally beyond the United States as part of its broader rollout (i.e., beyond an initial single‑city debut)?" It's a bullshit machine. 89%!
At some point you gotta face the music, right?
[0]: https://artificialanalysis.ai/evaluations/omniscience?model-...
[1]: https://huggingface.co/datasets/ArtificialAnalysis/AA-Omnisc...
You had to go all the way and find it in the benchmark results that specifically stress test this.
You could not come up with a single one yourself. And you also linked an example where it was not allowed to use tools when I specifically said that it should be able to use tools. I'm not sure why are you present this as though it is a big gotcha.
I think my main point pretty much stands.
I found over 500 examples that fit your criteria. Embarrassing you were arguing in bad faith this whole time.
They all use the tool search, no? Please correct me if I'm wrong.
My criteria was using ChatGPT which explicitly allows it.
https://arxiv.org/html/2511.13029v1 if you don't believe me.
BTW this was your original point
>Anyway, it's trivial to get pretty much any model to make things up. Don't we all know this? That's why I was surprised by your position; if we know anything about these things it's that they make things up.
And look at how much effort you have had to do
1. use the wrong model for the horns example
2. the game one also didn't work
3. now you are searching for examples in literal benchmarks and you are still not able to find any
How is this trivial in any interpretation of the word?
I think it would be perfectly reasonable to agree that it is not at all trivial to find counter examples for my challenge.
I've got about 20 minutes in this; mostly I've been reading wallstreetbets at the Shake Shack bar in the Boston airport. I'm happy to post this over and over again until you engage w/ it:
> I found over 500 examples that fit your criteria.
Oh, well you should have said that then.
You're talking to a different person there, but I do obviously also disagree with a lot of what's written in the post too.
At the same time, it is also just super redundant nevertheless, yes. Not sure why you find it so bizarre that one would take an issue with that. See also the very existence of the website called TV-Tropes.
Yeah, it gets really boring. Whenever I see "slot machines" or "bullshit machines" or whatever, I just ignore the comment and move on, because it signals that it's someone in such deep denial that they've turned their brain off.
I'd much rather read articles about what LLMs can/can't do, or stuff people have built with LLMs, than read how everything LLMs touch turns to shit.
Its usual gibberish that tries to throw many darts and see what sticks. Oh LLM's steal other people's work? Check. Oh LLM's cause ecological damage? Check. Oh LLM's hallucinate? Check.
When you see a pattern like this, you know that its not coming from any place of truth but rather ideology
My personal red flag for this is the scare quoting of AI, and the super try-hard categorization work that people perform to try and discredit LLMs.
It takes approximately 1 min to find out that machine learning is a subfield of artificial intelligence, both having existed for about half a century now. This basic historical fact is also taught on AI 101 courses across the globe for compsci students.
Yet here we are, people portraying it as some sort of cheap sales trick. Reminds me when I discussed quantum dots with a friend, which he was very enthusiastic to quickly file under "yet another bullshit with quantum in its name" before finally taking the time to understand that the "quantum" bit is not a marketing gimmick. Except in this case, people are a million times more inclined to willfully propagate this. Genuinely so tiresome.
I think it’s just anxiety because to internalise that it is actually so good is a bit hard for some