I really hope Anthropic turns out to be one of the 'good guys', or at least a net positive.
It appears they trend in the right direction:
- Have not kissed the Ring.
- Oppose blocking AI regulation that other's support (e.g. They do not support banning state AI laws [2]).
- Committing to no ads.
- Willing to risk defense department contract over objections to use for lethal operations [1]
The things that are concerning:
- Palantir partnership (I'm unclear about what this actually is) [3]
- Have shifted stances as competition increased (e.g. seeking authoritarian investors [4])
It inevitable that they will have to compromise on values as competition increases and I struggle parsing the difference marketing and actually caring about values. If an organization cares about values, it's suboptimal not to highlight that at every point via marketing. The commitment to no ads is obviously good PR but if it comes from a place of values, it's a win-win.
I'm curious, how do others here think about Anthropic?
Being the 'good guy' is just marketing. It's like a unique selling point for them. Even their name alludes to it. They will only keep it up as long as it benefits them. Just look at the comments from their CEO about taking Saudi money.
Not that I've got some sort of hate for Anthropic. Claude has been my tool of choice for a while, but I trust them about as much as I trust OpenAI.
How do you parse the difference between marketing and having values? I have difficulty with that and I would love to understand how people can be confident one way or the other. In many instances, the marketing becomes so disconnected from actions that it's obvious. That hasn't happen with Anthropic for me.
I am a fairly cynical person. Anthropic could have made this statement at any time, but they chose to do it when OpenAI says they are going to start showing ads, so view it in that context. They are saying this to try to get people angry about ads to drop OpenAI and move to Anthropic. For them, not having ads supports their current objective.
When you accept the amount of investments that these companies have, you don't get to guide your company based on principles. Can you imagine someone in a boardroom saying, "Everyone, we can't do this. Sure it will make us a ton of money, but it's wrong!" Don't forget, OpenAI had a lot of public goodwill in the beginning as well. Whatever principles Dario Amodei has as an individual, I'm sure he can show us with his personal fortune.
Parsing it is all about intention. If someone drops coffee on your computer, should you be angry? It depends on if they did it on purpose, or it was an accident. When a company posts a statement that ads are incongruous to their mission, what is their intention behind the message?
> Anthropic could have made this statement at any time, but they chose to do it when OpenAI says they are going to start showing ads, so view it in that context.
Obviously they did do it for that reason, but it does make sense. They've positioned themselves from day 1 as the AI company built on more values; that doesn't make them good but it's self-consistent. If, out of the blue earlier on when nobody was talking about ads in AI, they said "we're not going to put ads in AI", that would have been a Suspiciously Specific Denial: "our shirt saying we're not going to put ads in AI has people asking a lot of questions already answered by our shirt".
> Can you imagine someone in a boardroom saying, "Everyone, we can't do this. Sure it will make us a ton of money, but it's wrong!"
Yes. But that's not how you'd say it. "First of all, this would go against our established ethical principles, which you knew when you invested with us. Second, those ethical principles define our position in the market, which we should not abandon."
Ideally, ethical buyers would cause the market to line up behind ethical products. For that to be possible, we have to have choices available to us. Seems to me Anthropic is making such a choice available to see if buyers will line up behind it.
Companies, not begin sentient, don't have values, only their leaders/employees do. The question then becomes "when are the humans free to implement their values in their work, and when aren't they". You need to inspecting ownership structure, size, corporate charter and so on, and realize that it varies with time and situation.
My point is that the leaders have constraints on them that prevent them actually executing on their values. E.g. imagine leadership dislikes spam, but an institutional investor on the board has warned the CEO that if there's a sales dip before quarterly earnings and the market reacts badly he'll get fired. So the CEO - against his values - orders the VP or marketing to spam for all his life is worth. This stuff gets so internalized, that we routinely make decisions at work that go against our values because we know that's what's demanded of us by our organizations.
I think there are two key imperatives that lead to company "psychopathy".
The first imperative is a company must survive past its employees. A company is an explicit legal structure designed to survive past the initial people in the company. A company is _not_ the employees, it is what survives past the employees' employment.
The second imperative is the diffusion of responsibility. A company becomes the responsible party for actions taken, not individual employees. This is part of the reason we allow companies to survive past employees, because their obligations survive as well.
This leads to individual employees taking actions for the company against their own moral code for the good of the company.
See also The Corporation (2003 film) and Meditations On Moloch (2014)[0].
By providing product or services of value, not by maximing profits at any cost (definitely not by taking advantage of people, shortcomings of rules/laws, ... , or by harming people, ... , environment)
No company has values. Anthropic's resistance to the administration is only as strong as their incentive to resist, and that incentive is money. Their execs love the "Twitter vs Facebook" comparison that makes Sam Altman look so evil and gives them a relative halo effect. To an extent, Sam Altman revels in the evil persona that makes him appear like the Darth Vader of some amorphous emergent technology. Both are very profitable optics to their respective audiences.
If you lend any amount of real-world credence to the value of marketing, you're already giving the ad what it wants. This is (partially) why so many businesses pivoted to viral marketing and Twitter/X outreach that feels genuine, but requires only basic rhetorical comprehension to appease your audience. "Here at WhatsApp, we care deeply about human rights!" *audience loudly cheers*
Anthropic is a PBC, not a "company", and the people who work there basically all belong to AI safety as a religion. Being incredibly cynical is generally dumb, but it's especially dumb to apply "for profit company" incentives to something that isn't a traditional "for profit company".
The difference is if they are willingly ready to lose money on a personal level. If folks in the company are not willing to sacrifice their comp for good, they are not "good" guys.
For Anthropic and lot of startups with very high growth(even including OpenAI 4 years back or Google or Amazon), they don't have to lose anything to be good as they can just raise money. But when the growth stops that's when the test starts.
I believe in "too big to have values". No company that has grown beyond a certain size has ever had true values. Only shareholder wealth maximisation goals.
Anthropic is a PBC. The shareholder goals are public benefit (PB) not "wealth maximization".
(Also, wealth maximization is a dumb goal and not how successful companies work. Cynicism is a bad strategy for being rich because it's too shortsighted.)
Yes and OpenAI was a not for profit and look how that’s going. Now it’s a PBC. So Anthropic won’t even be the first PBC AI company pretending that they’re doing it for the good of the world and then trying to shove in porn and Add for wealth maximisation. Also most companies that go big have an IPO, and it’s mostly just about short term strategies to make share price go up after that.
What evidence do you have for that? Your point about Saudi is literally mentioned by the parent as one of the few negative points.
I’m not saying this is how it will play out, but this reads as lazy cynicism - which is a self-realising attitude and something I really don’t admire about our nerd culture. We should be aiming higher.
Agreed. Companies don’t have the capacity to be moral entities. They are driven purely based on monetary incentives. They are mechanical machinery. People are anthropomorphizing values onto companies or being duped by marketing speak.
I mean, yes and. Companies may do things for broadly marketing reasons, but that can have positive consequences for users and companies can make committed decisions that don't just optimize for short term benefits like revenue or share price. For example, Apple's commitment to user privacy is "just marketing" in a sense, but it does benefit users and they do sacrifice sources of revenue for it and even get into conflicts with governments over the issue.
And company execs can hold strong principles and act to push companies in a certain direction because of them, although they are always acting within a set of constraints and conflicting incentives in the corporate environment and maybe not able to impose their direction as far as they would like. Anthropic's CEO in particular seems unusually thoughtful and principled by the standards of tech companies, although of course as you say even he may be pushed to take money from unsavory sources.
Basically it's complicated. 'Good guys' and 'bad guys' are for Marvel movies. We live in a messy world and nobody is pure and independent once they are enmeshed within a corporate structure (or really, any strong social structure). I think we all know this, I'm not saying you don't! But it's useful to spell it out.
And I agree with you that we shouldn't really trust any corporations. Incentives shift. Leadership changes. Companies get acquired. Look out for yourself and try not to tie yourself too closely to anyone's product or ecosystem if it's not open source.
> and even get into conflicts with governments over the issue.
To be fair, they also cooperate with the US government for immoral dragnet surveillance[0], and regularly assent to censorship (VPN bans, removed emojis, etc.) abroad. It's in both Apple and most governments' best interests to appear like mortal enemies, but cooperate for financial and domestic security purposes. Which for all intents and purposes, it seems they do. Two weeks after the San Bernardino kerfuffle, the iPhone in question was cracked and both parties got to walk away conveniently vindicated of suspicion. I don't think this is a moral failing of anyone, it's just the obvious incentives of Apple's relationship with their domestic fed. Nobody holds Apple's morality accountable, and I bet they're quite grateful for that.
At the end of the day, the choices in companies we interact with is pretty limited. I much prefer to interact with a company that at least pays lip service to being 'good' as opposed to a company that is actively just plain evil and ok with it.
That's the main reason I stick with iOS. At least Apple talks about caring about privacy. Google/Android doesn't even bother to talk about it.
That's probably not true - government regulators require a lot of privacy work and Android certainly complies with that. Legal compliance is a large business strategy because small companies can't afford to do it.
They are the most anti-opensource AI Weights company on the planet, they don't want to do it and don't want anyone else to do it. They just hide behind safety and alignment blanket saying no models are safe outside of theirs, they wont even release their decommissioned models. Its just money play - Companies don't have ethics , the policies change based on money and who runs it - look at google - their mantra once was Don't be Evil.
Also codex cli, Gemini cli is open source - Claude code will never be - it’s their moat even though 100% written by ai as the creator says it never will be . Their model is you can use ours be it model or Claude code but don’t ever try to replicate it.
The steelman argument is that super-intelligent AGI could allow any random person to build destructive technology, so companies on the path toward creating that ought to be very careful about alignment, safety and, indeed, access to weights.
The obvious assumed premise of this argument is that Anthropic are actually on the path toward creating super-intelligent AGI. Many people, including myself, are skeptical of this. (In fact I would go farther - in my opinion, cosplaying as though their AI is so intelligent that it's dangerous has become a marketing campaign for Anthropic, and their rhetoric around this topic should usually be taken with a grain of salt.)
I would not consider myself an expert on LLMs, at least not compared to the people who actually create them at companies like Anthropic, but I can have a go at a steelman:
LLMs allow hostile actors to do wide-scale damage to society by significantly decreasing the marginal cost and increasing the ease of spreading misinformation, propaganda, and other fake content. While this was already possible before, it required creating large troll farms of real people, semi-specialized skills like photoshop, etc. I personally don't believe that AGI/ASI is possible through LLMs, but if you do that would magnify the potential damage tenfold.
Closed-weight LLMs can be controlled to prevent or at least reduce the harmful actions they are used for. Even if you don't trust Anthropic to do this alone, they are a large company beholden to the law and the government can audit their performance. A criminal or hostile nation state downloading an open weight LLM is not going to care about the law.
This would not be a particularly novel idea - a similar reality is already true of other products and services that can be used to do widespread harm. Google "Invention Secrecy Act".
I wouldn't mind doing my best steelman of the open source AI if he responds (seriously, id try).
Also, your comment is a bit presumptuous. I think society has been way too accepting of relying on services behind an online API, and it usually does not benefit the consumer.
I just think it's really dumb that people argue passionately about open weight LLMs without even mentioning the risks.
Since you asked for it, here is my steelman argument :
Everything can cause harm - it depends on who is holding it , how determined are they , how easy is it and what are the consequences. Open source will make this super easy and cheap.
1. We are already seeing AI Slop everywhere Social media Content, Fake Impersonation - if the revenue from whats made is larger than cost of making it , this is bound to happen, Open models can be run locally with no control, mostly it can be fine tuned to cause damage - where as closed source is hard as vendors might block it.
2. Less skilled person can exploit or create harmful code - who otherwise could not have.
3. Remove Guards from a open model and jailbreak, which can't be observed anymore (like a unknown zero day attack) since it may be running private.
4. Almost anything digital can be Faked/Manipulated from Original/Overwhelmed with false narratives so they can rank better over real in search.
The problem is that "good" companies cannot succeed in a landscape filled with morally bad ones, when you are in a time of low morality being rewarded. Competing in a rigged market by trying to be 100% morally and ethically right ends up in not competing at all. So companies have to pick and choose the hills they fight on. If you take a look at how people are voting with their dollars by paying for these tools...being a "good" company doesn't seem to factor much into it on aggregate.
exactly. you cant compete morally when cheating, doing illegal things and supporting bad guys are norm. Hence, I hope open models will win in the long term.
Similar to Oracle vs Postgres, or some closed source obscure caching vs Redis. One day I hope we will have very good SOTA open models where closed models compete to catch up (not saying Oracle is playing a catch up with Pg).
No good companies for you, yet you bet on Chinese labs! Even if you have no moral problems at all with the China authoritarian, Chinese companies are as morally trustworthy as American ones. That is clear.
As it’s often said: there is no such thing as free product, you are the product. AI training is expensive even for Chinese companies.
I expect to some degree the Chinese models don't need immediate profits, because having them as a show of capability for the state is already a goal met? They're probably getting some support from the state at least.
> Even if you have no moral problems at all with the China authoritarian
It's funny how you framed your sentence. Let's unpack it.
1. I didn't say Chinese companies are good, I said my hope is on open models and only Chinese labs are doing good in that front
2. Chinese company doesn't immediately mean its about regime. Maybe its true in the US with the current admin, see how Meta, Google, Microsoft got immediately aligned with current admin
3. Even when company is associated with Chinese regime, I don't remember Chinese authoritarian regime kidnapping the head of another state, invading bunch of countries in the Middle East and supporting states committing genocide and ethnic cleansing (Israel in Gaza, UAE in Sudan and many more small militant groups across Africa and ME), and authoritarian regimes like Saudi Arabia.
If you ask me to rate them by evil level, I would give the US 80/100, and China 25/100 - no invasions, no kidnapping of head of states, no obvious terror acts - but unfortunate situation with Uyghurs.
They are the only AI company more closed than OpenAI, which is quite a feat. Any "commitment" they make should only be interpreted as marketing until they rectify this. The only "good guys" in AI are the ones developing inference engines that let you run models on your own hardware. Any individual model has some problems, but by making models fungible and fully under the users control (access to weights) it becomes a possible positive force for the user.
>I really hope Anthropic turns out to be one of the 'good guys', or at least a net positive.
There are no good guys, Anthropic is one of the worst of the AI companies. Their CEO is continuously threatening all of the white collar workers, they have engineering playing the 100x engineer game on Xitter. They work with Palantir and support ICE. If anything, chinese companies are ethically better at this point.
I apologize for resurrecting a fairly old thread (I originally came across it via Googling the Claude Ad announcement), and this has nothing to do with the main post anyway but my desire not to further derail things is butting up against my pigheadedness as famously exemplified in xkcd.com/386
You wrote "the majority of US citizens support ICE", later specifying it was after the controversial killings. I want to fact-check your claim, because if I came here from Google I think others may too and so far there has been no pushback on this apart from "no, they don't support ICE" which is problematic.
The Navigator Research poll is probably the most interesting in terms of actually visualizing the data [5], it used to be at the bottom but I moved this section to the top. It's an excellent visualization, my eye was especially drawn to the incredibly stark divide between the MAGA-Republicans vs. non-MAGA Republicans. I've never seen it in such contrast and I wonder how that will play out. Regardless, that's outside of the topic of this comment, wich is already too long.
The poll (n=~1,000) conducted around the five following days states that among all those polled, the Net Favorable rating, asked as:
Please indicate how favorable or unfavorable you are to: ICE, also known as Immigration and Customs Enforcement.
was:
June 30 '25 - 8
Dec 8 '25 -14
Jan 12 '26 -20
Feb 1 '26 -22
As an added curio, among Democrats it shifted -12 points, but among Republicans it shifted -17 points. Make of that what you will.
A poll released on February 5th and conducted January 27th-30th (n=1,462) by Marist University's MaristPoll organization states [1]:
62% of Americans say the actions of ICE are making Americans somewhat less safe or much less safe. 37% think the actions of ICE are making Americans much more safe or somewhat more safe
The following information is from Pages 14-16 of the PDF [2]. The specific question was:
Do you approve or disapprove of the job Immigration and Customs Enforcement (ICE) is doing?
33% Approve
60% Disapprove
7% Unsure
The poll is extremely along the party lines in some respects (e.g. only 2% of the people who answered 'Approve' voted for Kamala in 2024), but I feel like that's to be expected. At the same time, while people who approve of ICE definitely didn't vote for Kamala, people who voted for Trump are not extremely strong supporters, with 24% of 2024 Trump voters answering either Disapprove or Unsure (18% and 6% respectively).
Some other interesting correlations are that while recent ICE actions are generally very unpopular among Democrats and somewhat popular among Republicans, the usual rural/urban differences which are usually much better predictors for how someone will vote, far ahead of correlations with education, race or income. In this case "approve" difference between Big city (31%) and Rural (43%) was only 12 points. Interestingly, the approval rating almost follows a reverse bell curve (?) in terms of population density:
Big city 31%
Small city 31%
Suburban 28%
Small town 40%
Rural 43%
Again, apologies for the digression, I figured maybe someone can make some sense of those numbers. Another question:
How would you describe the actions of Immigration and Customs Enforcement (ICE) in enforcing immigration laws? Do you think they have:
65% Gone too far
12% Not gone far enough
22% Or are their actions about right
1% Vol: unsure
This can be summarized in a bunch of different ways, but it's not dishonest to say that 2/3rds of Americans think that ICE has "gone too far", which heavily contradicts your statement. Everything after this is kind of just because I felt three is the minimum number of sources to thoroughly dismiss your assertion.
Here are the results from June, 2025 for the exact same question, with the change in percentage from June, 2025 to January, 2026 in brackets.
54% Gone too far [+11%]
18% Not gone far enough [-06%]
26% Or are their actions about right [-04%]
1% Vol: Unsure
Ipsos released a poll [3] on January 27th, conducted between January 23rd-25th (n=1,139) with a self-reported margin of error of +/- 3.0 percentage points at the 95% confidence level.
Immigration-related outtakes:
Only 10% of those polled picked "Immigration" as their selection to a 17-option question that asked "In your opinion, what is the most important problem facing the U.S. today?" By comparison, 15% picked the "Economy, unemployment and jobs", 12% picked "Political violence and extremism" and 16% picked "Threats to democratic values and norms" (Page 1)
Overall, do you think efforts by U.S. Immigration and Customs Enforcement (ICE) agents to deal with unauthorized immigration in the country... (Page 10)
Go too far 58%
Do not go far enough 12%
Are about right 26%
Do you approve or disapprove of the way President Donald Trump is handling the following issues? Immigration (Page 3)
Strongly approve 26%
Somewhat approve 14%
Somewhat disapprove 10%
Strongly disapprove 43%
Generally speaking, would you say the following things are heading in the right direction, or are they off on the wrong track? Immigration policy (Page 5)
Right direction 35%
Wrong track 51%
Don’t know 11%
P.S. I don't claim this comment to be unbiased, I clearly "cherry-picked" certain facts, but they were not out of context nor did I massage the numbers, the reason I phrased it this way was because it was a direct rebuttal to the claim that most Americans support ICE [even after the shootings].
I don’t know about “good guys” but the fact that they seem to be highly focused on coding rather than general purpose chat bot (hard to overcome chatGPT mindshare there) they have a customer base that is more willing to pay for usage and therefore are less likely to need to add an ad revenue stream. So yes so far I would say they are on stronger ground than the others.
When powerful people, companies, and other organizations like governments do a whole lot of very good and very bad things, figuring out whether this rounds to “more good than bad” or “more bad than good” is kind of a fraught question. I think Anthropic is still in the “more good than bad” range, but it doesn’t make sense to think about it along the lines of heros versus villains. They’ve done things that I put in the “seems bad” column, and will likely do more. Also more good things, too.
They’re moving towards becoming load-bearing infrastructure and then answering specific questions about what you should do about it become rather situational.
No one who believes this should be in any position of authority in the AI space. Anthropic's marketing BS has basically been taken as fact on this website since they started and it's just so tiring to watch this industry fall for the same nonsense over and over and over again.
Anthropic is younger. That's why they're not doing ads. As soon as they actually reach the spending to (not) reach their AGI goals they will start running ads and begging the taxpayer for even more money.
I always find the CEO using hype and fear as his marketing strategy.
A year ago, he came out and said, "Blood-bath" for white collar jobs.
It seems to create some sense of anxiety in the receiving end.
> I'm curious, how do others here think about Anthropic?
I’m very pleased they exist and have this mindset and are also so good at what they do. I have a Max subscription - my most expensive subscription by a wide margin - and don’t resent the price at all. I am earnestly and perhaps naively hoping they can avoid enshittification. A business model where I am not the product gives me hope.
Anthropic was founded by OpenAI defectors who said OpenAI's product strategy was too dangerous and needed more safety research. But in reality Anthropic has almost exactly the same product strategy. A lot of this is just marketing to raise money to make the founders billionaires rather than the multi-millionaires they only would've been if they hadn't founded a competitor.
In Poland, before the last presidential election, a member of one candidate’s campaign team had a moment of accidental honesty. Asked whether his candidate would pledge not to raise taxes after winning, he replied: “Well, what’s the harm in promising?”
I can't see how Google turned to become evil or how OpenAI did for that matter.
Google delivered on their promise, and OpenAI well it's too soon but it's looking good.
The name OpenAI and its structure is a relic from a world where the sentiment was to be heavily preoccupied and concerned by the potential accidental release of an AGI.
Now that it's time for products the name and the structure are no longer serving the goal
Their move of disallowing alternative clients to use a Claude Code subscription pissed me off immensely. I triggered a discussion about it yesterday[0]. It’s the opposite of the openness that led software to where it is today. I’m usually not so bothered about such things, but this is existential for us engineers. We need to scrutinise this behaviour from AI companies extra hard or we’re going to experience unprecedented enshittification. Imagine a world where you’ve lost your software freedoms and have no ability to fight back because Anthropic’s customers are pumping out 20x as many features as you.
Anthropic's move of disallowing opencode is quite offputting to me because there really isn't a way to interpret it as anything other than a walled-garden move that abuses their market position to deliberately lock in users.
Opencode ought to have similar usage patterns to Claude Code, being a very similar software (if anything Opencode would use fewer tokens as it doesn't have some fancy features from Claude Code like plan files and background agents). Any subscription usage pattern "abuses" that you can do with Opencode can also be done by running Claude Code automatically from the CLI. Therefore restricting Opencode wouldn't really save Anthropic money as it would just move problem users from automatically calling Opencode to automatically calling CC. The move seems to purely be one to restrict subscribers from using competing tools and enforce a vertically-integrated ecosystem.
In fact, their competitor OpenAI has already realized that Opencode is not really dissimilar from other coding agents, which is why they are comfortable officially supporting Opencode with their subscription in the first place. Since Codex is already open-source and people can hack it however they want, there's no real downside for OpenAI to support other coding agents (other than lock-in). The users enter through a different platform, use the service reasonably (spending a similar amount of tokens as they would with Codex), and OpenAI makes profit from these users as well as PR brownie points for supporting an open ecosystem.
In my mind being in control of the tools I use is a big feature when choosing an AI subscription and ecosystem to invest into. By restricting Opencode, Anthropic has managed to turn me off from their product offerings significantly, and they've managed to do so even though I was not even using Opencode. I don't care about losing access to a tool I'm not using, but I do care about what Anthropic signals with this move. Even if it isn't the intention to lock us in and then enshittify the product later, they are certainly acting like it.
The thing is, I am usually a vote-with-my-wallet person who would support Anthropic for its values even if they fall behind significantly compared to competitors. Now, unless they reverse course on banning open-source AI tools, I will probably revert to simply choosing whichever AI company is ahead at any given point.
I don't know whether Anthropic knows that they are pissing off their most loyal fanbase of conscientious consumers a lot with these moves. Sure, we care about AI ethics and safety, but we also care about being treated well as consumers.
What year do you think it is? The US is actively aggressive in multiple areas of the world. As a non US citizen I don’t think helping that effort at the expense of the rest of the world is good.
Yeah. That's all absurd. It's a big leap from that to the conclusion that the US military isn't a net good for the world. What post-modern world do you think you're living in in which militaries aren't needed for defense? And if they weren't around to defend you, you'd be better off? The veneer of civilization isn't as thick as you seem to think.
I mean that it's one thing to think it's bad to be defending the US. And it's another thing to think it's bad for the US to be defending the place you live.
I'd prefer we kicked out the usians and reformed back into people's defense forces.
I've spent some time in the military but didn't pursue it because they made a law where we could be ordered to participate in foreign missions, and at the time I knew people who had been to Afghanistan and helped out protecting US drug barons. They weren't feeling very good and got very little assistance with their psychiatric issues, like going to parties and drinking and crying a lot and sometimes tell stories about kids they'd murdered.
That sounds reasonable but is it realistic to believe that with military spending of 10% of your potential adversary (i.e., Russia) Sweden would be able to mount a credible defense on its own?
> - Willing to risk defense department contract over objections to use for lethal operations [1]
> The things that are concerning: - Palantir partnership (I'm unclear about what this actually is) [3]
Dude, you cannot put these two sentences together. The defense department was either a fluke or a PR stunt. If they partner with Palintir they absolutely do not care that their tech is going to be used for killing and other horrible deeds.
A company with morals (which does not exist BTW) would never partner with Palintir.
Given that LLMs essentially stole business models from public (and not!) works the ideal state is they all die in favor of something we can run locally.
I feel like they are picking a lane. ChatGPT is great for chatbots and the like, but, as was discussed in a prior thread, chatbots aren't the end-all-be-all of AI or LLMs. Claude Code is the workhorse for me and most folks I know for AI assisted development and business automation type tasks. Meanwhile, most folks I know who use ChatGPT are really replacing Google Search. This is where folks are trying to create llm.txt files to become more discoverable by ChatGPT specifically.
For Anthropic to be proactive in saying they will not pursue ad based revenue I think is not just "one of the good guys" but that they may be stabilizing on a business model of both seat and usage based subscriptions.
Either way, both companies are hemorrhaging money.
> ChatGPT is saying they will mark ads as ads and keep answers "independent," but that is not measurable. So we'll see.
Yeah I remember when Google used to be like this. Then today I tried to go to 39dollarglasses.com and accidentally went to the top search result which was actually an ad for some other company. Arrrg.
Before Google, web search was a toxic stew of conflicts of interest. It was impossible to tell if search results were paid ads or the best possible results for your query.
Google changed all that, and put a clear wall between organic results and ads. They consciously structured the company like a newspaper, to prevent the information side from being polluted and distorted by the money-making side.
Here's a snip from their IPO letter [0]:
Google users trust our systems to help them with important decisions: medical, financial and many others. Our search results are the best we know how to produce. They are unbiased and objective, and we do not accept payment for them or for inclusion or more frequent updating. We also display advertising, which we work hard to make relevant, and we label it clearly. This is similar to a well-run newspaper, where the advertisements are clear and the articles are not influenced by the advertisers’ payments. We believe it is important for everyone to have access to the best information and research, not only to the information people pay for you to see.
Anthropic's statement reads the same way, and it's refreshing to see them prioritize long-term values like trust over short-term monetization.
It's hard to put a dollar value on trust, but even when they fall short of their ideals, it's still a big differentiator from competitors like Microsoft, Meta and OpenAI.
I'd bet that a large portion of Google's enterprise value today can be traced to that trust differential with their competitors, and I wouldn't be surprised to see a similar outcome for Anthropic.
I agree. Having watched Google shift from its younger idealistic values to its current corrupted state, I can't help but be cynical about Anthropic's long-term trajectory.
But if nothing else, I can appreciate Anthropic's current values, and hope they will last as long as possible...
I use OpenCode and I made an "architect" agent that uses Opus to make a plan, then gives that plan to a "developer" agent (with Sonnet) that implements it, and a "reviewer" agent (Codex) reviews it in the end. I've gotten much better results with this than with straight up Opus throughout, and obviously hit the limits much less often as well.
Agreed on using both. I definitely know people who prefer Codex or Cursor. It's probably Coke or Pepsi at this point. I tend to prefer Claude Code, but that's just me.
You may not like this sources, but both the tomato throwers to the green visor crowds agree they are losing money. How and when they make up the difference is up to speculation
The comment was with reference to inference, not total P&L.
Of course they are losing money in total. They are not, however, losing money per marginal token.
It’s trivial to see this by looking at the market clearing price of advanced open source models and comparing to the inference prices charged by OpenAI.
The closed frontier models seem to sell at a substantial premium to inference on open-source models, so that does suggest that there is a decent margin to the inference. The training is where they're losing money, and the bull case is that every model makes money eventually, but the models keep getting bigger or at least more expensive to train, so they're borrowing money to make even more money later (which does need to converge somehow, i.e. they can't just keep shooting larger until the market can't actually afford to pay for the training). The bear case is that this is basically just a treadmill to stay on the frontier where they can make that premium (if the big labs ever stop they'll quickly get caught up by cheaper or even open-source models and lose their edge), in which case it's probably never going to actually become sustainable.
> If we subtract the cost of compute from revenue to calculate the gross margin (on an accounting basis),2 it seems to be about 50% — lower than the norm for software companies (where 60-80% is typical) but still higher than many industries.
This is one of those “don’t be evil” like articles that companies remove when the going gets tough but I guess we should be thankful that things are looking rosy enough for Anthropic at the moment that they would release a blog like this.
The point about filtering signal vs. noise in search engines can’t really be stated enough. At this point using a search engine and the conventional internet in general is an exercise in frustration. It’s simply a user hostile place – infinite cookie banners for sites that shouldn’t collect data at all, auto play advertisements, engagement farming, sites generated by AI to shill and produce a word count. You could argue that AI exacerbates this situation but you also have to agree that it is much more pleasant to ask perplexity, ChatGPT or Claude a question than to put yourself through the torture of conventional search. Introducing ads into this would completely deprive the user of a way of navigating the web in a way that actually respects their dignity.
I also agree in the sense that the current crop of AIs do feel like a space to think as opposed to a place where I am being manipulated, controlled or treated like some sheep in flock to be sheared for cash.
The current crop of LLM-backed chatbots do have a bit of that “old, good internet” flavor. A mostly unspoiled frontier where things are changing rapidly, potential seems unbounded, the people molding the actual tech and discussing it are enthusiasts with a sort of sorcerer’s apprentice vibe. Not sure how long it can persist, since I’ve seen this story before and we all understand the incentive structures at play.
Does anyone know how if there are precedents for PBCs or B-Corp type businesses to be held accountable for betraying their stated values? Or is it just window dressing with no legal clout? Can they change to a standard corporation on a whim and ditch the non-shareholder maximization goals?
There’s nothing old internet about these AI companies. Old internet was about giving out and asking for nothing in return. These companies take everything and give back nothing, unless you are willing to pay that is.
I get the sentiment, but if you can't acknowledge that AI is useful and currently a lot better than search for a great many things, then it's hard to have a rational conversation.
No, they don't. They soak up tons of your most personal and sensitive information like a sponge, and you don't know what's done with it. In the "good old Internet", that did not happen. Also in the good old Internet, it wasn't the masses all dependent on a few central mega-corporations shaping the interaction, but a many-to-many affair, with people and organizations of different sizes running the sites where interaction took place.
Ok, I know I'm describing the past with rosy glasses. After all, the Internet started as a DARPA project. But still, current reality is itself rather dystopic in many ways.
> This is one of those “don’t be evil” like articles that companies remove when the going gets tough but I guess we should be thankful that things are looking rosy enough for Anthropic at the moment that they would release a blog like this.
Exactly this. Show me the incentive, and I'll show you the outcome, but at least I'm glad we're getting a bit more time ad-free.
Right, if there's no legal weight to any of their statements then they mean almost nothing. It's a very weak signal and just feels like marketing. All digital goods can and will be made worse over time if it benefits the company.
Current LLMs often produce much, much worse results than manually searching.
If you need to search the internet on a topic that is full of unknown unknowns for you, they're a pretty decent way to get a lay of the land, but beyond that, off to Kagi (or Google) you go.
Even worse is that the results are inconsistent. I can ask Gemini five times at what temperature I should take a waterfowl out of the oven, and get five different answers, 10°C apart.
What do you mean, "are you sure"? I literally saw and see it happen in front of my eyes. Just now tested it with slight variations of "ideal temperature waterfowl cooking", "best temperature waterfowl roasting", etc. and all these questions yield different answers, with temperatures ranging from 47c-57c (ignoring the 74c food safety ones).
That's my entire point. Even adding an "is" or "the" can get you way different advice. No human would give you different info when you ask "what's the waterfowl's best cooking temperature" vs "what is waterfowl's best roasting temperature".
Did you point that out to one of them… like “hey bro, I’ve asked y’all this question in multiple threads and get wildly different answers. Why?”
And the answer is probably because there is no such thing as an ideal temperature for waterfowl because the answer is “it depends” and you didn’t give it enough context to better answer your question.
Context is everything. Give it poor prompts, you’ll get poor answers. LLMs are no different than programming a computer or anything else in this domain.
And learning how to give good context is a skill. One we all need to learn.
But that isn't how normal people interact with search engines. Which is the whole argument everyone is saying here, how LLMs are now better 'correct answer generators' than search engine. They're not. My mother directly experienced that. Her food would have come out much better if she completely ignored Gemini and checked a site.
One of the best things LLMs could do (and that no one seems to be doing) is allow it to admit uncertainty. If the average weight of all tokens in a response drops below X, it should just say "I don't know, you should check a different source."
At any rate, if my mother has to figure out some 10 sentence stunted multiform question for the LLM to finally get a good consistent answer, or can just type "best Indian restaurant in Brooklyn" (maybe even with site:restaurant reviews.com"), which experience is superior?
> LLMs are no different than programming a computer or anything else in this domain.
Just feel like reiterating against this: virtually no one programs their search queries or query engineers a 10 sentence search query.
If I made a new, not-AI tool called 'correct answer provider' which provided definitive, incorrect answers to things you'd call it bad software. But because it is AI we're going to blame the user for not second guessing the answers or holding it wrong ie. bad prompting.
I created an account just to point out that this is simply not true. I just tried it! The answers were consistent across all 5 samples with both "Fast" mode and Pro (which I think is really important to mention if you're going to post comments like this - I was thinking maybe it would be inconsistent with the Flash model)
It obviously takes discipline, but using something like Perplexity as an aggregator typically gets me better results, because I can click through to the sources.
It's not a perfect solution because you need the discipline/intuition to do that, and not blindly trust the summary.
My mother did, for Christmas. It was a goose that ended up being raw in a lot of places.
I then pointed out this same inconsistency to her, and that she shouldn't put stock in what Gemini says. Testing it myself, it would give results between 47c-57c. And sometimes it would just trip out and give the health-approved temperature, which is 74c (!).
Edit: just tested it again and it still happens. But inconsistency isn't a surprise for anyone who actually knows how LLMs work.
> But inconsistency isn't a surprise for anyone who actually knows how LLMs work
Exactly. These people saying they've gotten good results for the same question aren't countering your argument. All they're doing is proving that sometimes it can output good results. But a tool that's randomly right or wrong is not a very useful one. You can't trust any of its output unless you can validate it. And for a lot of the questions people ask of it, if you have to validate it, there was no reason to use the LLM in the first place.
I'm only 25 and grew up in a post "don't be evil" world so maybe I'm jaded, but it always seems strange to me when companies try to make any moral stand about anything. Presumably if there was something a company could do to make more money they'd do it regardless of the ethical standing behind it b/c
1. The people running the company aren't stupid and want to make more money
2. The company is beholden to investors to make more money both practically and legally
IMO this means the only reason anthropic isn't adding ads is b/c they make more money posturing as the "good guys". As soon as that's no longer true not only are the people in charge going to do it, but there's probably some legal standing that would obligate/compel them to do so
A company is just a group of people. Anything a group of people can believe, a company can believe.
A company with investors is a little more suspect, its true. But even massive public companies like Costco still stake out some claims with values beyond the economic and then pursue them. It's not an impossibility.
I also think, despite dropping the motto, that Google did not in fact ever become anything close to evil. I think people are being somewhat unfair (and hyperbolic at best) by claiming otherwise.
Does it matter if they are truly moral or truly cynical but think they can capture some sort of "morality premium" by taking a stance?
Yes they are beholden to investors. But so is Musk. The thing is that his investors are also his fans, so a majority of them will approve whatever tactics or schemes he comes up with.
Anthropic may be able to cultivate investors in a similar way, without even trying. After all, why would you invest in a company if you didn't agree with how it was run? Of course there could be a hostile takeover by private equity group who thinks they can squeeze more value out of the tech, but that is harder the larger the company grows.
Edit: also, the morality premium is usually enjoyed more in terms of employees than investors. Google for a long time attracted idealists, by which I mean they were able to pay slightly below market rate for a certain kind of person. If you have a moral compass and you get an equal offer from oai and anthropic right now, which one you choosing?
The key hurdle for AI to leap is establishing trust with users. No one trusts the big players (for good reason) and it is causing serious anxiety among the investors. It seems Claude acknowledges this and is looking to make trust a critical part of their marketing messaging by saying no ads or product placement. The problem is that serving ads is only one facet of trust. There are trust issues around privacy, intellectual property, transparency, training data, security, accuracy, and simply "being evil" that Claude's marketing doesn't acknowledge or address. Trust, on the scale they need, is going to be very hard for any of them to establish, if not impossible.
That still doesn't mean much unless you're doing your own training or getting the weights from a trusted source, and neither of those mean much without knowing something about the data being trained on.
If someone is trying to influence your results, running the inference on your own infrastructure prevents some attack vectors but not some of the more plausible and worrying ones.
I don't think people are concerned about the models' math being biased/tainted (people know of it but that largely doesn't factor into the "security concerns" that people cite.) Typically, it's about how do we know that our data is not going to be seen by a 3rd party. That's what I'm speaking to. Running on your own infra, you can guarantee there are no phone-homes.
What do you mean? Google is roughly the most trusted organization in the world by revealed preference. The 800(?) million ChatGPT users – I have a hard time reading that as a trust problem.
Usage metrics don't reveal preference in all cases. The fact these companies are sketchy/untrustworthy is practically a meme, including among non-tech people. Their services are widely relied upon, but they enjoy very little subjective good will
I always found Anthropic to be trying hard to signal as one of the "good guys".
I wonder how they can get away without showing Ads when ChatGPT has to be doing it. Will the enterprise business be that profitable that Ads are not required?
Maybe OpenAI is going for something different - democratising access to vast majority of the people. Remember that ChatGPT is what people know about and what people use the free version of.
Who's to say that making Ads by doing this but also prodiding more access is the wrong choice?
Also, Claude holds nothing against ChatGPT in search. From my previous experiences, ChatGPT is just way better at deep searches through the internet than Claude.
ChatGPT is providing a ridiculous amount of free service to gain/keep traction. Others also have free tiers, but to a much lesser extent. It's similar to Uber selling rides at a loss to win markets. It will get you traction, yes, but the bill has to be paid one day.
Even when you're subscribed, they're providing unreasonable amounts of compute for the price. I am subscribed to both Claude and ChatGPT, and Claude's limits are so tiny compared to ChatGPT's that it often feels like a rip-off.
None of the ai companies are, they are all looking for those multi billion deals to provide the backplane for services like Copilot and Siri.
Consumer chatbots are pure marketing, no company is going to make anything off those $20 per month subs to ai chatbots.
Besides the editorial control -which openai openly flagged to want to remain unbiased- there is a deeper issue with ads-based revenue models in AI: that of margins. If you want ads to cover compute & make margins -looking at roughly $50 ARPU at mature FB/GOOG level- you have two levers: sell more advertisement, or offer dumber models.
This is exactly what chatgpt 5 was about. By tweaking both the model selector (thinking/non-thinking), and using a significantly sparser thinking model (capping max spend per conversation turn), they massively controlled costs, but did so at the expense of intelligence, responsiveness, curiosity, skills, and all the things I've valued in O3. This was the point I dumped openai, and went with claude.
This business model issue is a subtle one, but a key reason why advertisement revenue model is not compatible (or competitive!) with "getting the best mental tools" -margin-maximization selects against businesses optimizing for intelligence.
The vast majority of people don't need smarter models and aren't willing to pay for a subscription. There's an argument to be made that ads on free users will subsidize the power users that demand frontier intelligence - done well this could increase OpenAI's revenue by an order of magnitude.
This is going to be tough to compete against - Anthropic would need to go stratospheric with their (low margin) enterprise revenue.
This will be an amusing post to revisit in the internet archives when or if they do introduce ads in the future but dressed up in a different presentation and naming. Ultimately the investors will come calling.
They are using this to virtue signal - but in reality it's just not compatible with their businesses model.
Anthropic is mainly focusing on B2B/Enterprise and tool use cases, in terms of active users I'd guess Claude is distant last, but in terms of enterprise/paying customers I wouldn't be surprised if they were ahead of the others.
I believe Perplexity is doing this already, but specifically for looking up products, which is how I use AI sometimes. I am wondering how long before eBay, Amazon etc partner with AI companies to give them more direct API access so they can show suggested products and what not. I like how AI can summarize things for me when looking up products, then I open up the page and confirm for myself.
Won't all the ad revenue come from commerce use cases ... and they seem to be excluding that from this announcement:
> AI will increasingly interact with commerce, and we look forward to supporting this in ways that help our users. We’re particularly interested in the potential of agentic commerce
Why bother with ads when you can just pay an AI platform to prefer products directly? Then every time an agentic decision occurs, the product preference is baked in, no human in the loop. AdTech will be supplanted by BriberyTech.
The only chance of that happening is if Altman somehow feels sufficiently shamed into abandoning the lazy enshittification track to monetization.
I don't think they have an accurate model for what they're doing - they're treating it like just another app or platform, using tools and methods designed around social media and app store analytics. They're not treating it like what it is, which is a completely novel technology with more potential than the industrial revolution for completely reshaping how humans interact with each other and the universe, fundamentally disrupting cognitive labor and access to information.
The total mismatch between what they're doing with it to monetize and what the thing actually means to civilization is the biggest signal yet that Altman might not be the right guy to run things. He's savvy and crafty and extraordinarily good at the palace intrigue and corporate maneuvering, but if AdTech is where they landed, it doesn't seem like he's got the right mental map for AI, for all he talks a good game.
There are a number of different llms - no reason they all need to do things the same. If you are replacing web search then ads are probably how you earn money. However if you are replacing the work people do for a company it makes more sense to charge for the work. I'm not sure if their current token charges are the right one, but it seems like a better track.
yeah it’s either that or openai has effected a massive own-goal… im leaning toward your view, but hoping that prediction does not manifest. i would be fine with all sorts of shit in life being more expensive but ad-free… but this is certainly a priviledged take and i recognize that.
> Anthropic is focused on businesses, developers, and helping our users flourish. Our business model is straightforward: we generate revenue through enterprise contracts and paid subscriptions, and we reinvest that revenue into improving Claude for our users. This is a choice with tradeoffs, and we respect that other AI companies might reasonably reach different conclusions.
Very diplomatic of them to say "we respect that other AI companies might reasonably reach different conclusions" while also taking a dig at OpenAI on their youtube channel
I think this says a lot about the business approach of Anthopic compared to OpenAI. Just the vast amount of free messages you get from OpenAI is crazy that turning a profit with that seems impossible. Anthropic is growing more slowly but it seems like they are not running a crazy deficit. They do not need to put ads or porn in their chatbot
"Think Different" also used to mean something, before Tim Cook cozied up to authoritarianism. Who knows, it's never too late for Anthropic to have a change of heart.
They are not trying to sell adds. They are trying to sell themselves as a
monthly service. That is what I think when they are trying to convince me to go there to think. I rather go think at Wikipedia.
Idk, brainstorming and ideating is my main use case for AI
I use it as codegen too but I easily have 20x more brainstorming conversations than code projects
Most non-tech people I talk to are finding value with it with traditional things. The main one I've seen flourish is travel planning. Like, booking became super easy but full itinerary planning for a trip (hotels, restaurants, day trips/activities, etc) has been largely a manual thing that I see a lot of non-tech people using llms for. It's very good for open ended plans too, which the travel sites have been horrible at. For instance, "I want to plan a trip to somewhere warm and beachy I don't care about the dates or exactly where" maybe I care about the budget up front but most things I'm flexible on - those kinds of things work well as a conversation.
Wikipedia is, of course very useful, but what it’s not good at is surfacing information I am unfamiliar with. Part of this problem is that Wikipedia editors are more similar to me, and more interested in similar things to me, than the average person writing text that appears online. Part of the problem is that the design of Wikipedia does not make it easy to stumble upon unexpected information; most links are to adjacent topics given they have to be relevant to the current article. But regardless, I’m much more likely to come across a novel concept when chatting with Claude, compared to browsing Wikipedia.
It’s so hard to succeed without selling ads. There’s an exponential growth aspect to these endeavors and ads add a lot of revenue, which investors like, so those who don’t can find that the lost revenue “multiplies” due to lower outside investment, lower stock price growth, etc.
I wish the financial aspects were different, because Anthropic is absolutely correct about ads being antithetical to a good user experience.
Anthropic is very big (the biggest AI co?) in B2B, where you don't have ads. Also, if they end up creating a datacenter full of geniuses, ads won't make sense either.
Sometimes posts like this are just value-signaling. I hear a lot of cynicism and "just you wait, the other shoe will drop" comments along those lines.
But combined with the other projects Anthropic has pursued (e.g. around understanding bias and explaining "how the model is thinking as it is") and decisions it has made, I'm happy with the course they're plotting. They seem consistently upstanding, thoughtful, and respectful. I want to commend them and earnestly say: Keep up the good work!
> "I'm very interested in the idea of a piece of music being a place to think," Richter explained, adding that he had written Daylight as a response to the 2003 Iraq War. [0]
Seems short-sighted to commit to not running ads exactly like OpenAI plans to. You win in the court of public opinion for a few months, then look like a hypocrite when you're inevitably forced to run ads because your investors demand it. Sort of like how safetyism was convenient marketing until it became clear that it was revenue repellant and they quietly walked it away.
I asked for this last week in an hn comment and people were pretty negative about it in the replies.
But I’m happy with position and will cancel my ChatGPT and push my family towards Claude for most things. This taste effect is what I think pushes apple devices into households. Power users making endorsements.
And I think that excess margin is enough to get past lowered ad revenue opportunity.
So apparently they're going to run a Super Bowl ad about ChatGPT having ads (without saying ChatGPT of course)........ Has doing an ad that focuses only on something about your competitor ever been the best play? Talk about yourself.
Obviously it's a play, honing in on privacy/anti-ad concerns, like a Mozilla type angle, but really it's a huge ad buy just to slag off the competitors. Worth the expense just to drive that narrative?
That's a pretty weird take. Her costume is typical of track and field athletes of the time, except that her jersey represents the nation of the Apple Macintosh, which you can barely notice in the ad because it flashes by in a split second.
ah, good one. Was it Big Blue or Big Brother in general being referenced in that one? Either way I suppose Apple didn't even say much of anything about their product in that one where Anthropic is at least highlighting a feature.
> There are many good places for advertising. A conversation with Claude is not one of them.
> ...but including ads in conversations with Claude would be incompatible with what we want Claude to be: a genuinely helpful assistant for work and for deep thinking.
Sadly, with my disillusionment with the tech industry, plus the trend of the past 20 years, this smacks of Larry Page's early statements about how bad advertising could distort search results and Google would never do that. Unsurprisingly, I am not able to find the exact quote with Google.
I appreciate taking a stance, even if nobody is asking. It would be great if it was less of a bad faith effort.
It's great that Anthropic is targeting the businesses of the world. It's a little insincere to than declare "no ads", as if that decision would obviously be the same if the bulk of their (not paying) users.
There are, as far as ads go, perfectly fine opportunities to do them in a limited way for limited things within chatbots. I don't know who they think they are helping by highlighting how to do it poorly.
Not the point of the article, but just responding to the headline: LLMs are a space to think in the same way that watching television is a space to think. Some people may think while watching TV, maybe many of the people reading this watch TV this way, but most will not. In fact, television mostly acts as a soporific, killing intelligent thought and numbing the mind against any sort of complex cognition. Same with LLMs. They're a space to sleepily watch someone else think.
So they have "made a choice" to keep Claude ad-free, they say. "Today [...] Claude’s only incentive is to give a helpful answer", they say. But there's nothing that suggests that they can't make a different choice tomorrow, or whenever it suits them. It's not profitable to betray your trust too early.
I can't really imagine any statement they could give that would ease concerns that at some point in time they change their mind. But for now, it is a relief to read, even if this is a bit of marketing. The longer it goes without being enshittified the better.
They could agree to some actual significant negative consequence to running ads. e.g. They could put a clause in the subscription signup process that says if they ever run ads - even if it's only for free accounts - then you get all of the money you've spent back.
Of course I realise they would never do something like that. Buy why not? Well, because they might decide they want to run ads...
Ben Thompson has long been insistent that ChatGPT and other AI tools basically have to have ads and it's been a big mistake they didn't have them sooner. It's an interesting take:
> What I think is clear is they have to build an advertising product, and the reason they have to build an advertising product is any consumer Internet product has to be advertising, because it’s such a beneficial model to everyone involved, and the reason it’s so beneficial is you get to indefinitely and infinitely increase average revenue per user without any worries about price elasticity, because the entire increase in average revenue per user is borne by the advertisers who are paying it willingly because they’re getting a positive return on their investment, and everyone’s using it for free so you can reach the whole world. Then what happens with that is once you get that model going, you have a massive R&D advantage, because you have so much more money coming in than anyone who doesn’t have that cycle or who has to charge users for it.
> This point, more than anything else, explains why the company so desperately needs an advertising model. Advertising is the only potential business model that can meaningfully bend the revenue curve such that the company can not just fund its compute but gain leverage on it, for all of the reasons I laid out before: first, advertising increases the breadth of the business, in that you can offer a better product to more people, increasing usage and expanding inventory. Second, advertising increases the depth of the business, in that there is infinite upside in terms of average revenue per user: more usage means more inventory on one hand, and building out the capability for effective targeting and high conversion rates increases the amount that advertisers are willing to pay — even as the cost to the user remains the same (ideally free).
It's valuable to remember that advertisers will pay more per user than users will, and that's hard to beat in a competitive market.
100%. Love this approach by Anthropic. The Meta "monetization league" is assembling at OpenAI and doing what they've done best at Meta.
However, I do think we need to take Anthropic's word with a grain of salt, too. To say they're fully working in the user's interest has yet to be proven. This trust would require a lot of effort to be earned. Once the companies intends to or becomes public, incentives change, investors expect money and throwing your users under the bus is a tried and tested way of increasing shareholder value.
I have such mixed feelings about Anthropic. Opus 4.5 is legit, Claude Code has been making solid progress (but it's gonna get stomped by pi) and they've been trying to be more transparent. On the other hand, they still infra issues, they till haven't open sourced Claude Code and they're going after other harnesses that use their sub. Dario also seems to have a very paternalistic attitude that I don't like, though I don't think that will survive the market.
I really want to applaud Anthropic; I remain cautiously optimistic, but I’m not certain how long they will maintain this posture. I will say that the recent announcement from OpenAI has put me off from ChatGPT — I use Gemini occasionally, because it’s the devil I know. OpenAI has gone back and forth on their positions so many times in a way that feels truly hostile to their users.
What makes Anthropic seem like early Apple is not just the unique taste, but the courage to stand firm with their vision of what the product should be.
I am going to assume that is a question asked in good faith by an inquisitive mind. That assumed, asking that question to any proper LLM will provide an answer more in depth than I am willing to type this evening.
Palantir, to me, is the weaponization of big data, where advanced analytics are used to target vulnerable populations. Not just abroad, but here against its own citizens. It is the dystopic enabler that we have been warned about.
Palantir and the words from its leadership seem to me to be in direct opposition to parts of the Constitution doc that Anthropic hold up to show their ethics and seriousness.
>An advertising-based business model would introduce incentives that could work against this principle.
I agree with this - I'm not so much worried that ChatGPT is going to silently insert advertising copy into model answers. I'm worried that advertising alongside answers creates bad incentives that then drive future model development. We saw Google Search go down this path.
Using ads in AI, today, is like printing flyers when the Internet started, (or sending email ads 30 years ago: of course Google could have promised: "we will not send spam to your gmail mailbox..." :).
"Ads" aim to influence your behavior: AI is a doing much much more with no need for ads (Claude included)
Claude have posted on number of very sarcastic videos on twitter that take a jibe at ads https://x.com/claudeai/status/2019071118036942999 with an ending line "Ads are coming to IA. But not to Claude."
I do feel like many folks here and elsewhere are missing the mark with LLMs.
1) Yes, they are absolutely useless in a consumer setting.
2) If you want to be a software developer, you absolutely need to know how to understand/interact with one, and you more than likely will need to understand things like https://continue.dev.
I am no longer in software development due to my body slowly (quickly) dying, however I see it all from the sidelines:
1) New tech was rushed to the front lines way too quickly by big tech.
2) Big (and small tech) rushed layoffs way too fast rather than let we devs explore the advantages vs. disadvantages.
3) Companies blame "AI" (LLMs) for layoffs.
4) Most senior devs (including myself) soundly reject AI due to the above.
5) New generation of devs uses AI tools, some struggle occurs where morons don't bother reviewing code that was written by an auto completion engine.
6) We nerds begin to understand the usefulness of LLMs for "the boring part"
Not a shareholder of any company. I'm permanently disabled. Just watching this stuff from the sidelines.
What other interaction models exist for Claude given that Anthropic seems to be stressing so much that this is for "conversations"?
(Props for them for doing this, don't know how this is long-term sustainable for them though ... especially given they want to IPO and there will be huge revenue/margin pressures)
Anthropic probably saw how much money they made off of the Moltbot hype and figured that they don’t need ad revenue. They can go a step further and build a marketplace for similar setups, paying the developers who make them in micro transactions per tokens.
I actually do want ads. Subtle, labeled ads. For many queries in Google, I like to know who is advertising for an intent. In may cases, that is a perfect fit. I don’t want ONLY the companies who have gamed the LLM SEO market to win.
Good on Anthropic! I appreciate how deliberate they are on maintaining user trust. Have preferred Claude's responses more through the API, so I don't imagine this would have affected me as much but it is still nice to see.
I would object to ads across the board in this case (though I’m generally fine with even targeted ads). It would create a customer-client relationship between companies paying to advertise and the AI company, creating an incentive for Anthropic to manipulate the Claude service on their behalf. As an end user that seeks input from Claude on purchasing decisions, I do not want there to be any question as to whether or not it was subtly manipulated.
It's nice that they don't show ads in conversations with Claude - but I wonder if they collect profiling information from my prompts and activities to sell to advertising firms.
Who do they think believe the whole "don't be evil" in 2026?
We know what's around the corner. Enshitification, loss of trust, frog boiling, account restrictions and upsell, advertising, degradation of service, data sold for advertising and worse.
Y'ain't kidding anyone with this stuff. You're only providing screenshots for future memes.
If you broach subjects Anthropic considers sensitive (cyber security, dangerous biotech, etc) Claude is very likely to shut you down completely and refuse to answer. As someone that works in cybersecurity and uses Claude daily, it is annoying to ask a question regarding some feature of Cobalt Strike and have it refuse to answer, even though the tool’s documentation is public. I would have cancelled my ChatGPT subscription at this point if once or twice a month I didn’t need to ask it to look up something when Claude refuses.
I agree that there are a ton of tells, but I wouldn't say picking up some new vocabulary is "speaking like" just yet. It's still early, though, so more studies are needed.
I really hope Anthropic turns out to be one of the 'good guys', or at least a net positive.
It appears they trend in the right direction:
- Have not kissed the Ring.
- Oppose blocking AI regulation that other's support (e.g. They do not support banning state AI laws [2]).
- Committing to no ads.
- Willing to risk defense department contract over objections to use for lethal operations [1]
The things that are concerning: - Palantir partnership (I'm unclear about what this actually is) [3]
- Have shifted stances as competition increased (e.g. seeking authoritarian investors [4])
It inevitable that they will have to compromise on values as competition increases and I struggle parsing the difference marketing and actually caring about values. If an organization cares about values, it's suboptimal not to highlight that at every point via marketing. The commitment to no ads is obviously good PR but if it comes from a place of values, it's a win-win.
I'm curious, how do others here think about Anthropic?
[1]https://archive.is/Pm2QS
[2]https://www.nytimes.com/2025/06/05/opinion/anthropic-ceo-reg...
[3]https://investors.palantir.com/news-details/2024/Anthropic-a...
[4]https://archive.is/4NGBE
Being the 'good guy' is just marketing. It's like a unique selling point for them. Even their name alludes to it. They will only keep it up as long as it benefits them. Just look at the comments from their CEO about taking Saudi money.
Not that I've got some sort of hate for Anthropic. Claude has been my tool of choice for a while, but I trust them about as much as I trust OpenAI.
How do you parse the difference between marketing and having values? I have difficulty with that and I would love to understand how people can be confident one way or the other. In many instances, the marketing becomes so disconnected from actions that it's obvious. That hasn't happen with Anthropic for me.
I am a fairly cynical person. Anthropic could have made this statement at any time, but they chose to do it when OpenAI says they are going to start showing ads, so view it in that context. They are saying this to try to get people angry about ads to drop OpenAI and move to Anthropic. For them, not having ads supports their current objective.
When you accept the amount of investments that these companies have, you don't get to guide your company based on principles. Can you imagine someone in a boardroom saying, "Everyone, we can't do this. Sure it will make us a ton of money, but it's wrong!" Don't forget, OpenAI had a lot of public goodwill in the beginning as well. Whatever principles Dario Amodei has as an individual, I'm sure he can show us with his personal fortune.
Parsing it is all about intention. If someone drops coffee on your computer, should you be angry? It depends on if they did it on purpose, or it was an accident. When a company posts a statement that ads are incongruous to their mission, what is their intention behind the message?
> Anthropic could have made this statement at any time, but they chose to do it when OpenAI says they are going to start showing ads, so view it in that context.
Obviously they did do it for that reason, but it does make sense. They've positioned themselves from day 1 as the AI company built on more values; that doesn't make them good but it's self-consistent. If, out of the blue earlier on when nobody was talking about ads in AI, they said "we're not going to put ads in AI", that would have been a Suspiciously Specific Denial: "our shirt saying we're not going to put ads in AI has people asking a lot of questions already answered by our shirt".
> Can you imagine someone in a boardroom saying, "Everyone, we can't do this. Sure it will make us a ton of money, but it's wrong!"
Yes. But that's not how you'd say it. "First of all, this would go against our established ethical principles, which you knew when you invested with us. Second, those ethical principles define our position in the market, which we should not abandon."
Ideally, ethical buyers would cause the market to line up behind ethical products. For that to be possible, we have to have choices available to us. Seems to me Anthropic is making such a choice available to see if buyers will line up behind it.
“Ideally” is doing a lot of heavy lifting here.
Wow. Well said.
Companies, not begin sentient, don't have values, only their leaders/employees do. The question then becomes "when are the humans free to implement their values in their work, and when aren't they". You need to inspecting ownership structure, size, corporate charter and so on, and realize that it varies with time and situation.
Anthropic being a PBC probably helps.
>Companies, not begin sentient, don't have values, only their leaders/employees do
Isn't that a distinction without a difference? Every real world company has employees, and those people do have values (well, except the psychopaths).
My point is that the leaders have constraints on them that prevent them actually executing on their values. E.g. imagine leadership dislikes spam, but an institutional investor on the board has warned the CEO that if there's a sales dip before quarterly earnings and the market reacts badly he'll get fired. So the CEO - against his values - orders the VP or marketing to spam for all his life is worth. This stuff gets so internalized, that we routinely make decisions at work that go against our values because we know that's what's demanded of us by our organizations.
I think there are two key imperatives that lead to company "psychopathy".
The first imperative is a company must survive past its employees. A company is an explicit legal structure designed to survive past the initial people in the company. A company is _not_ the employees, it is what survives past the employees' employment.
The second imperative is the diffusion of responsibility. A company becomes the responsible party for actions taken, not individual employees. This is part of the reason we allow companies to survive past employees, because their obligations survive as well.
This leads to individual employees taking actions for the company against their own moral code for the good of the company.
See also The Corporation (2003 film) and Meditations On Moloch (2014)[0].
[0] https://slatestarcodex.com/2014/07/30/meditations-on-moloch/
> How do you parse the difference between marketing and having values?
You don't. Companies want people to think they have values. But companies are not people. Companies exist to earn money.
> That hasn't happen with Anthropic for me.
Yet.
> Companies exist to earn money
By providing product or services of value, not by maximing profits at any cost (definitely not by taking advantage of people, shortcomings of rules/laws, ... , or by harming people, ... , environment)
Not sure if you're being sarcastic or not
Humans don't have values either
what the heck heck
No company has values. Anthropic's resistance to the administration is only as strong as their incentive to resist, and that incentive is money. Their execs love the "Twitter vs Facebook" comparison that makes Sam Altman look so evil and gives them a relative halo effect. To an extent, Sam Altman revels in the evil persona that makes him appear like the Darth Vader of some amorphous emergent technology. Both are very profitable optics to their respective audiences.
If you lend any amount of real-world credence to the value of marketing, you're already giving the ad what it wants. This is (partially) why so many businesses pivoted to viral marketing and Twitter/X outreach that feels genuine, but requires only basic rhetorical comprehension to appease your audience. "Here at WhatsApp, we care deeply about human rights!" *audience loudly cheers*
Anthropic is a PBC, not a "company", and the people who work there basically all belong to AI safety as a religion. Being incredibly cynical is generally dumb, but it's especially dumb to apply "for profit company" incentives to something that isn't a traditional "for profit company".
The difference is if they are willingly ready to lose money on a personal level. If folks in the company are not willing to sacrifice their comp for good, they are not "good" guys.
For Anthropic and lot of startups with very high growth(even including OpenAI 4 years back or Google or Amazon), they don't have to lose anything to be good as they can just raise money. But when the growth stops that's when the test starts.
People have values, Corporations do not.
I believe in "too big to have values". No company that has grown beyond a certain size has ever had true values. Only shareholder wealth maximisation goals.
Anthropic is a PBC. The shareholder goals are public benefit (PB) not "wealth maximization".
(Also, wealth maximization is a dumb goal and not how successful companies work. Cynicism is a bad strategy for being rich because it's too shortsighted.)
Yes and OpenAI was a not for profit and look how that’s going. Now it’s a PBC. So Anthropic won’t even be the first PBC AI company pretending that they’re doing it for the good of the world and then trying to shove in porn and Add for wealth maximisation. Also most companies that go big have an IPO, and it’s mostly just about short term strategies to make share price go up after that.
What evidence do you have for that? Your point about Saudi is literally mentioned by the parent as one of the few negative points.
I’m not saying this is how it will play out, but this reads as lazy cynicism - which is a self-realising attitude and something I really don’t admire about our nerd culture. We should be aiming higher.
While there is nothing wrong about Saudi money, they are spreading FUD about open weights and wants to regulate and control how the LLM is used
Regulation is not always bad. What’s the FUD you refer to?
Agreed. Companies don’t have the capacity to be moral entities. They are driven purely based on monetary incentives. They are mechanical machinery. People are anthropomorphizing values onto companies or being duped by marketing speak.
> but I trust them about as much as I trust OpenAI.
So, ideally, not at all?
I mean, yes and. Companies may do things for broadly marketing reasons, but that can have positive consequences for users and companies can make committed decisions that don't just optimize for short term benefits like revenue or share price. For example, Apple's commitment to user privacy is "just marketing" in a sense, but it does benefit users and they do sacrifice sources of revenue for it and even get into conflicts with governments over the issue.
And company execs can hold strong principles and act to push companies in a certain direction because of them, although they are always acting within a set of constraints and conflicting incentives in the corporate environment and maybe not able to impose their direction as far as they would like. Anthropic's CEO in particular seems unusually thoughtful and principled by the standards of tech companies, although of course as you say even he may be pushed to take money from unsavory sources.
Basically it's complicated. 'Good guys' and 'bad guys' are for Marvel movies. We live in a messy world and nobody is pure and independent once they are enmeshed within a corporate structure (or really, any strong social structure). I think we all know this, I'm not saying you don't! But it's useful to spell it out.
And I agree with you that we shouldn't really trust any corporations. Incentives shift. Leadership changes. Companies get acquired. Look out for yourself and try not to tie yourself too closely to anyone's product or ecosystem if it's not open source.
> and even get into conflicts with governments over the issue.
To be fair, they also cooperate with the US government for immoral dragnet surveillance[0], and regularly assent to censorship (VPN bans, removed emojis, etc.) abroad. It's in both Apple and most governments' best interests to appear like mortal enemies, but cooperate for financial and domestic security purposes. Which for all intents and purposes, it seems they do. Two weeks after the San Bernardino kerfuffle, the iPhone in question was cracked and both parties got to walk away conveniently vindicated of suspicion. I don't think this is a moral failing of anyone, it's just the obvious incentives of Apple's relationship with their domestic fed. Nobody holds Apple's morality accountable, and I bet they're quite grateful for that.
[0] https://arstechnica.com/tech-policy/2023/12/apple-admits-to-...
At the end of the day, the choices in companies we interact with is pretty limited. I much prefer to interact with a company that at least pays lip service to being 'good' as opposed to a company that is actively just plain evil and ok with it.
That's the main reason I stick with iOS. At least Apple talks about caring about privacy. Google/Android doesn't even bother to talk about it.
That's probably not true - government regulators require a lot of privacy work and Android certainly complies with that. Legal compliance is a large business strategy because small companies can't afford to do it.
They are the most anti-opensource AI Weights company on the planet, they don't want to do it and don't want anyone else to do it. They just hide behind safety and alignment blanket saying no models are safe outside of theirs, they wont even release their decommissioned models. Its just money play - Companies don't have ethics , the policies change based on money and who runs it - look at google - their mantra once was Don't be Evil.
https://www.anthropic.com/news/anthropic-s-recommendations-o...
Also codex cli, Gemini cli is open source - Claude code will never be - it’s their moat even though 100% written by ai as the creator says it never will be . Their model is you can use ours be it model or Claude code but don’t ever try to replicate it.
They don't even want people using OpenCode with their Max subscriptions (which OpenAI does allow, kind of)
[flagged]
Think of the harm that bad actors could cause with access to high school chemistry and physics textbooks
The steelman argument is that super-intelligent AGI could allow any random person to build destructive technology, so companies on the path toward creating that ought to be very careful about alignment, safety and, indeed, access to weights.
The obvious assumed premise of this argument is that Anthropic are actually on the path toward creating super-intelligent AGI. Many people, including myself, are skeptical of this. (In fact I would go farther - in my opinion, cosplaying as though their AI is so intelligent that it's dangerous has become a marketing campaign for Anthropic, and their rhetoric around this topic should usually be taken with a grain of salt.)
I would not consider myself an expert on LLMs, at least not compared to the people who actually create them at companies like Anthropic, but I can have a go at a steelman:
LLMs allow hostile actors to do wide-scale damage to society by significantly decreasing the marginal cost and increasing the ease of spreading misinformation, propaganda, and other fake content. While this was already possible before, it required creating large troll farms of real people, semi-specialized skills like photoshop, etc. I personally don't believe that AGI/ASI is possible through LLMs, but if you do that would magnify the potential damage tenfold.
Closed-weight LLMs can be controlled to prevent or at least reduce the harmful actions they are used for. Even if you don't trust Anthropic to do this alone, they are a large company beholden to the law and the government can audit their performance. A criminal or hostile nation state downloading an open weight LLM is not going to care about the law.
This would not be a particularly novel idea - a similar reality is already true of other products and services that can be used to do widespread harm. Google "Invention Secrecy Act".
"please do all the work to argue my position so I don't have to".
I wouldn't mind doing my best steelman of the open source AI if he responds (seriously, id try).
Also, your comment is a bit presumptuous. I think society has been way too accepting of relying on services behind an online API, and it usually does not benefit the consumer.
I just think it's really dumb that people argue passionately about open weight LLMs without even mentioning the risks.
Since you asked for it, here is my steelman argument : Everything can cause harm - it depends on who is holding it , how determined are they , how easy is it and what are the consequences. Open source will make this super easy and cheap. 1. We are already seeing AI Slop everywhere Social media Content, Fake Impersonation - if the revenue from whats made is larger than cost of making it , this is bound to happen, Open models can be run locally with no control, mostly it can be fine tuned to cause damage - where as closed source is hard as vendors might block it. 2. Less skilled person can exploit or create harmful code - who otherwise could not have. 3. Remove Guards from a open model and jailbreak, which can't be observed anymore (like a unknown zero day attack) since it may be running private. 4. Almost anything digital can be Faked/Manipulated from Original/Overwhelmed with false narratives so they can rank better over real in search.
I am on the opposite side of what you are thinking.
- Blocking access to others (cursor, openai, opencode)
- Asking to regulate hardware chips more, so that they don't get good competition from Chinese labs
- partnerships with palantir, DoD as if it wasn't obvious how these organizations use technology and for what purposes.
at this scale, I don't think there are good companies. My hope is on open models, and only labs doing good in that front are Chinese labs.
The problem is that "good" companies cannot succeed in a landscape filled with morally bad ones, when you are in a time of low morality being rewarded. Competing in a rigged market by trying to be 100% morally and ethically right ends up in not competing at all. So companies have to pick and choose the hills they fight on. If you take a look at how people are voting with their dollars by paying for these tools...being a "good" company doesn't seem to factor much into it on aggregate.
exactly. you cant compete morally when cheating, doing illegal things and supporting bad guys are norm. Hence, I hope open models will win in the long term.
Similar to Oracle vs Postgres, or some closed source obscure caching vs Redis. One day I hope we will have very good SOTA open models where closed models compete to catch up (not saying Oracle is playing a catch up with Pg).
No good companies for you, yet you bet on Chinese labs! Even if you have no moral problems at all with the China authoritarian, Chinese companies are as morally trustworthy as American ones. That is clear.
As it’s often said: there is no such thing as free product, you are the product. AI training is expensive even for Chinese companies.
I expect to some degree the Chinese models don't need immediate profits, because having them as a show of capability for the state is already a goal met? They're probably getting some support from the state at least.
> Even if you have no moral problems at all with the China authoritarian
It's funny how you framed your sentence. Let's unpack it.
1. I didn't say Chinese companies are good, I said my hope is on open models and only Chinese labs are doing good in that front
2. Chinese company doesn't immediately mean its about regime. Maybe its true in the US with the current admin, see how Meta, Google, Microsoft got immediately aligned with current admin
3. Even when company is associated with Chinese regime, I don't remember Chinese authoritarian regime kidnapping the head of another state, invading bunch of countries in the Middle East and supporting states committing genocide and ethnic cleansing (Israel in Gaza, UAE in Sudan and many more small militant groups across Africa and ME), and authoritarian regimes like Saudi Arabia.
If you ask me to rate them by evil level, I would give the US 80/100, and China 25/100 - no invasions, no kidnapping of head of states, no obvious terror acts - but unfortunate situation with Uyghurs.
> Blocking access
> Asking to regulate hardware chips more
> partnerships with [the military-industrial complex]
> only labs doing good in that front are Chinese labs
That last one is a doozy.
I agree, they seem to be following the Apple playbook. Make a closed off platform and present yourself as morally superior.
They are the only AI company more closed than OpenAI, which is quite a feat. Any "commitment" they make should only be interpreted as marketing until they rectify this. The only "good guys" in AI are the ones developing inference engines that let you run models on your own hardware. Any individual model has some problems, but by making models fungible and fully under the users control (access to weights) it becomes a possible positive force for the user.
>I really hope Anthropic turns out to be one of the 'good guys', or at least a net positive.
There are no good guys, Anthropic is one of the worst of the AI companies. Their CEO is continuously threatening all of the white collar workers, they have engineering playing the 100x engineer game on Xitter. They work with Palantir and support ICE. If anything, chinese companies are ethically better at this point.
Even in CNN polling the majority of US citizens support ICE.
Perhaps your moral bubble is not universal.
Murder is immoral. That much is certain. ICE murders people.
No one's been murdered. And those polls were after the protesters trying to kill police were shot in self-defense.
I apologize for resurrecting a fairly old thread (I originally came across it via Googling the Claude Ad announcement), and this has nothing to do with the main post anyway but my desire not to further derail things is butting up against my pigheadedness as famously exemplified in xkcd.com/386
You wrote "the majority of US citizens support ICE", later specifying it was after the controversial killings. I want to fact-check your claim, because if I came here from Google I think others may too and so far there has been no pushback on this apart from "no, they don't support ICE" which is problematic.
The Navigator Research poll is probably the most interesting in terms of actually visualizing the data [5], it used to be at the bottom but I moved this section to the top. It's an excellent visualization, my eye was especially drawn to the incredibly stark divide between the MAGA-Republicans vs. non-MAGA Republicans. I've never seen it in such contrast and I wonder how that will play out. Regardless, that's outside of the topic of this comment, wich is already too long.
The poll (n=~1,000) conducted around the five following days states that among all those polled, the Net Favorable rating, asked as:
was: As an added curio, among Democrats it shifted -12 points, but among Republicans it shifted -17 points. Make of that what you will.A poll released on February 5th and conducted January 27th-30th (n=1,462) by Marist University's MaristPoll organization states [1]:
The following information is from Pages 14-16 of the PDF [2]. The specific question was: The poll is extremely along the party lines in some respects (e.g. only 2% of the people who answered 'Approve' voted for Kamala in 2024), but I feel like that's to be expected. At the same time, while people who approve of ICE definitely didn't vote for Kamala, people who voted for Trump are not extremely strong supporters, with 24% of 2024 Trump voters answering either Disapprove or Unsure (18% and 6% respectively).Some other interesting correlations are that while recent ICE actions are generally very unpopular among Democrats and somewhat popular among Republicans, the usual rural/urban differences which are usually much better predictors for how someone will vote, far ahead of correlations with education, race or income. In this case "approve" difference between Big city (31%) and Rural (43%) was only 12 points. Interestingly, the approval rating almost follows a reverse bell curve (?) in terms of population density:
Again, apologies for the digression, I figured maybe someone can make some sense of those numbers. Another question: This can be summarized in a bunch of different ways, but it's not dishonest to say that 2/3rds of Americans think that ICE has "gone too far", which heavily contradicts your statement. Everything after this is kind of just because I felt three is the minimum number of sources to thoroughly dismiss your assertion.Here are the results from June, 2025 for the exact same question, with the change in percentage from June, 2025 to January, 2026 in brackets.
Ipsos released a poll [3] on January 27th, conducted between January 23rd-25th (n=1,139) with a self-reported margin of error of +/- 3.0 percentage points at the 95% confidence level.Immigration-related outtakes:
Only 10% of those polled picked "Immigration" as their selection to a 17-option question that asked "In your opinion, what is the most important problem facing the U.S. today?" By comparison, 15% picked the "Economy, unemployment and jobs", 12% picked "Political violence and extremism" and 16% picked "Threats to democratic values and norms" (Page 1)
P.S. I don't claim this comment to be unbiased, I clearly "cherry-picked" certain facts, but they were not out of context nor did I massage the numbers, the reason I phrased it this way was because it was a direct rebuttal to the claim that most Americans support ICE [even after the shootings].[1] https://maristpoll.marist.edu/polls/the-actions-of-ice-febru... [2] https://maristpoll.marist.edu/wp-content/uploads/2026/02/NPR... [3] https://www.ipsos.com/en-us/majority-americans-say-federal-i... [4] https://www.ipsos.com/sites/default/files/ct/news/documents/... [5] https://navigatorresearch.org/wp-content/uploads/2026/02/Nav...
I don’t know about “good guys” but the fact that they seem to be highly focused on coding rather than general purpose chat bot (hard to overcome chatGPT mindshare there) they have a customer base that is more willing to pay for usage and therefore are less likely to need to add an ad revenue stream. So yes so far I would say they are on stronger ground than the others.
When powerful people, companies, and other organizations like governments do a whole lot of very good and very bad things, figuring out whether this rounds to “more good than bad” or “more bad than good” is kind of a fraught question. I think Anthropic is still in the “more good than bad” range, but it doesn’t make sense to think about it along the lines of heros versus villains. They’ve done things that I put in the “seems bad” column, and will likely do more. Also more good things, too.
They’re moving towards becoming load-bearing infrastructure and then answering specific questions about what you should do about it become rather situational.
They're 16% owned by Google and Amazon, so they're already a minimum of 16% "bad guys".
Remember when OpenAI was about not-for-profit AI development for the betterment of humanity?
> Committing to no ads.
No one who believes this should be in any position of authority in the AI space. Anthropic's marketing BS has basically been taken as fact on this website since they started and it's just so tiring to watch this industry fall for the same nonsense over and over and over again.
Anthropic is younger. That's why they're not doing ads. As soon as they actually reach the spending to (not) reach their AGI goals they will start running ads and begging the taxpayer for even more money.
I am not sure, if it will turn out to be.
I always find the CEO using hype and fear as his marketing strategy. A year ago, he came out and said, "Blood-bath" for white collar jobs. It seems to create some sense of anxiety in the receiving end.
> I'm curious, how do others here think about Anthropic?
I’m very pleased they exist and have this mindset and are also so good at what they do. I have a Max subscription - my most expensive subscription by a wide margin - and don’t resent the price at all. I am earnestly and perhaps naively hoping they can avoid enshittification. A business model where I am not the product gives me hope.
Anthropic was founded by OpenAI defectors who said OpenAI's product strategy was too dangerous and needed more safety research. But in reality Anthropic has almost exactly the same product strategy. A lot of this is just marketing to raise money to make the founders billionaires rather than the multi-millionaires they only would've been if they hadn't founded a competitor.
Anthropic hasn't released image or video generation models. Seems pretty different to me.
Claude is somewhat sycophantic but nowhere near 4o levels. (or even Gemini 3 levels)
In Poland, before the last presidential election, a member of one candidate’s campaign team had a moment of accidental honesty. Asked whether his candidate would pledge not to raise taxes after winning, he replied: “Well, what’s the harm in promising?”
Google was the 'good guy.' Until it isn't.
Hell, OpenAI was the good guy.
I can't see how Google turned to become evil or how OpenAI did for that matter.
Google delivered on their promise, and OpenAI well it's too soon but it's looking good.
The name OpenAI and its structure is a relic from a world where the sentiment was to be heavily preoccupied and concerned by the potential accidental release of an AGI.
Now that it's time for products the name and the structure are no longer serving the goal
On not having kissed the ring - the board appointments and nat sec appointments indicate to me this is not true, thoughts?
Their move of disallowing alternative clients to use a Claude Code subscription pissed me off immensely. I triggered a discussion about it yesterday[0]. It’s the opposite of the openness that led software to where it is today. I’m usually not so bothered about such things, but this is existential for us engineers. We need to scrutinise this behaviour from AI companies extra hard or we’re going to experience unprecedented enshittification. Imagine a world where you’ve lost your software freedoms and have no ability to fight back because Anthropic’s customers are pumping out 20x as many features as you.
[0]: https://news.ycombinator.com/item?id=46873708
Anthropic's move of disallowing opencode is quite offputting to me because there really isn't a way to interpret it as anything other than a walled-garden move that abuses their market position to deliberately lock in users.
Opencode ought to have similar usage patterns to Claude Code, being a very similar software (if anything Opencode would use fewer tokens as it doesn't have some fancy features from Claude Code like plan files and background agents). Any subscription usage pattern "abuses" that you can do with Opencode can also be done by running Claude Code automatically from the CLI. Therefore restricting Opencode wouldn't really save Anthropic money as it would just move problem users from automatically calling Opencode to automatically calling CC. The move seems to purely be one to restrict subscribers from using competing tools and enforce a vertically-integrated ecosystem.
In fact, their competitor OpenAI has already realized that Opencode is not really dissimilar from other coding agents, which is why they are comfortable officially supporting Opencode with their subscription in the first place. Since Codex is already open-source and people can hack it however they want, there's no real downside for OpenAI to support other coding agents (other than lock-in). The users enter through a different platform, use the service reasonably (spending a similar amount of tokens as they would with Codex), and OpenAI makes profit from these users as well as PR brownie points for supporting an open ecosystem.
In my mind being in control of the tools I use is a big feature when choosing an AI subscription and ecosystem to invest into. By restricting Opencode, Anthropic has managed to turn me off from their product offerings significantly, and they've managed to do so even though I was not even using Opencode. I don't care about losing access to a tool I'm not using, but I do care about what Anthropic signals with this move. Even if it isn't the intention to lock us in and then enshittify the product later, they are certainly acting like it.
The thing is, I am usually a vote-with-my-wallet person who would support Anthropic for its values even if they fall behind significantly compared to competitors. Now, unless they reverse course on banning open-source AI tools, I will probably revert to simply choosing whichever AI company is ahead at any given point.
I don't know whether Anthropic knows that they are pissing off their most loyal fanbase of conscientious consumers a lot with these moves. Sure, we care about AI ethics and safety, but we also care about being treated well as consumers.
Too late for that, they came out and said they will train on anything you type in there months ago
They work with the US military.
Defending the US. So?
What year do you think it is? The US is actively aggressive in multiple areas of the world. As a non US citizen I don’t think helping that effort at the expense of the rest of the world is good.
Two things can be true. The US pays for most of the defense of NATO.
You’re currently threatening to invade a NATO country and being investigated for electoral interference in another. Not buying it anymore.
Whatever is going on there, I would not categorise a company backing that as morally “good”
Yeah. That's all absurd. It's a big leap from that to the conclusion that the US military isn't a net good for the world. What post-modern world do you think you're living in in which militaries aren't needed for defense? And if they weren't around to defend you, you'd be better off? The veneer of civilization isn't as thick as you seem to think.
Sure, as well as other powers are actively aggressive during last N thousands years, that's how humans operate, who don't they extinct.
The US military is famous for purely acting in self defence...
That's pretty bad.
Sweden too. So there's that.
What do you mean?
I mean that it's one thing to think it's bad to be defending the US. And it's another thing to think it's bad for the US to be defending the place you live.
It doesn't. US presence where I live, make it a target.
M'okay. Cops are evil, military is evil. Looks bleak.
I'd prefer we kicked out the usians and reformed back into people's defense forces.
I've spent some time in the military but didn't pursue it because they made a law where we could be ordered to participate in foreign missions, and at the time I knew people who had been to Afghanistan and helped out protecting US drug barons. They weren't feeling very good and got very little assistance with their psychiatric issues, like going to parties and drinking and crying a lot and sometimes tell stories about kids they'd murdered.
That sounds reasonable but is it realistic to believe that with military spending of 10% of your potential adversary (i.e., Russia) Sweden would be able to mount a credible defense on its own?
> - Willing to risk defense department contract over objections to use for lethal operations [1]
> The things that are concerning: - Palantir partnership (I'm unclear about what this actually is) [3]
Dude, you cannot put these two sentences together. The defense department was either a fluke or a PR stunt. If they partner with Palintir they absolutely do not care that their tech is going to be used for killing and other horrible deeds.
A company with morals (which does not exist BTW) would never partner with Palintir.
I think I’m not allowed to say what I think should happen to anyone who works with Palantir.
Maybe you could use an LLM to clean up what you want to say
[dead]
Given that LLMs essentially stole business models from public (and not!) works the ideal state is they all die in favor of something we can run locally.
Anthropic settled with authors of stolen work for $1.5b, this case is closed, isn't it?
Its not approved yet I think.
[dead]
I feel like they are picking a lane. ChatGPT is great for chatbots and the like, but, as was discussed in a prior thread, chatbots aren't the end-all-be-all of AI or LLMs. Claude Code is the workhorse for me and most folks I know for AI assisted development and business automation type tasks. Meanwhile, most folks I know who use ChatGPT are really replacing Google Search. This is where folks are trying to create llm.txt files to become more discoverable by ChatGPT specifically.
You can see the very different response by OpenAI: https://openai.com/index/our-approach-to-advertising-and-exp.... ChatGPT is saying they will mark ads as ads and keep answers "independent," but that is not measurable. So we'll see.
For Anthropic to be proactive in saying they will not pursue ad based revenue I think is not just "one of the good guys" but that they may be stabilizing on a business model of both seat and usage based subscriptions.
Either way, both companies are hemorrhaging money.
> ChatGPT is saying they will mark ads as ads and keep answers "independent," but that is not measurable. So we'll see.
Yeah I remember when Google used to be like this. Then today I tried to go to 39dollarglasses.com and accidentally went to the top search result which was actually an ad for some other company. Arrrg.
Before Google, web search was a toxic stew of conflicts of interest. It was impossible to tell if search results were paid ads or the best possible results for your query.
Google changed all that, and put a clear wall between organic results and ads. They consciously structured the company like a newspaper, to prevent the information side from being polluted and distorted by the money-making side.
Here's a snip from their IPO letter [0]:
Google users trust our systems to help them with important decisions: medical, financial and many others. Our search results are the best we know how to produce. They are unbiased and objective, and we do not accept payment for them or for inclusion or more frequent updating. We also display advertising, which we work hard to make relevant, and we label it clearly. This is similar to a well-run newspaper, where the advertisements are clear and the articles are not influenced by the advertisers’ payments. We believe it is important for everyone to have access to the best information and research, not only to the information people pay for you to see.
Anthropic's statement reads the same way, and it's refreshing to see them prioritize long-term values like trust over short-term monetization.
It's hard to put a dollar value on trust, but even when they fall short of their ideals, it's still a big differentiator from competitors like Microsoft, Meta and OpenAI.
I'd bet that a large portion of Google's enterprise value today can be traced to that trust differential with their competitors, and I wouldn't be surprised to see a similar outcome for Anthropic.
Don't be evil, but unironically.
[0] https://abc.xyz/investor/founders-letters/ipo-letter/default...
I agree. Having watched Google shift from its younger idealistic values to its current corrupted state, I can't help but be cynical about Anthropic's long-term trajectory.
But if nothing else, I can appreciate Anthropic's current values, and hope they will last as long as possible...
Disagree.
I end up using ChatGPT for general coding tasks because of the limited session/weekly limit Claude pro offers, and it works surprisingly well.
The best is IMO to use them both. They complement each other.
I use OpenCode and I made an "architect" agent that uses Opus to make a plan, then gives that plan to a "developer" agent (with Sonnet) that implements it, and a "reviewer" agent (Codex) reviews it in the end. I've gotten much better results with this than with straight up Opus throughout, and obviously hit the limits much less often as well.
Agreed on using both. I definitely know people who prefer Codex or Cursor. It's probably Coke or Pepsi at this point. I tend to prefer Claude Code, but that's just me.
Both companies are making bank on inference
You may not like this sources, but both the tomato throwers to the green visor crowds agree they are losing money. How and when they make up the difference is up to speculation
https://www.wheresyoured.at/why-everybody-is-losing-money-on... https://www.economist.com/business/2025/12/29/openai-faces-a... https://finance.yahoo.com/news/openais-own-forecast-predicts...
The comment was with reference to inference, not total P&L.
Of course they are losing money in total. They are not, however, losing money per marginal token.
It’s trivial to see this by looking at the market clearing price of advanced open source models and comparing to the inference prices charged by OpenAI.
> green visor crowds
??
https://en.wikipedia.org/wiki/Green_eyeshade
That is the big question. Got reliable data on that?
(My gut feeling tells me Claude Code is currently underpriced with regards to inference costs. But that's just a gut feeling...)
https://www.wheresyoured.at/costs/
Their AWS spend being higher than their revenue might hint at the same.
Nobody has reliable data, I think it's fair to assume that even Anthropic is doing voodoo math to sleep at night.
The closed frontier models seem to sell at a substantial premium to inference on open-source models, so that does suggest that there is a decent margin to the inference. The training is where they're losing money, and the bull case is that every model makes money eventually, but the models keep getting bigger or at least more expensive to train, so they're borrowing money to make even more money later (which does need to converge somehow, i.e. they can't just keep shooting larger until the market can't actually afford to pay for the training). The bear case is that this is basically just a treadmill to stay on the frontier where they can make that premium (if the big labs ever stop they'll quickly get caught up by cheaper or even open-source models and lose their edge), in which case it's probably never going to actually become sustainable.
> If we subtract the cost of compute from revenue to calculate the gross margin (on an accounting basis),2 it seems to be about 50% — lower than the norm for software companies (where 60-80% is typical) but still higher than many industries.
https://epoch.ai/gradient-updates/can-ai-companies-become-pr...
The context of that quote is OpenAI as a whole.
Maybe on the API, but I highly doubt that the coding agent subscription plans are profitable at the moment.
Build out distribution first and generate network effects.
For sure not
Could you substantiate that? That take into account training and staffing costs?
The parent specifically said inference, which does not include training and staffing costs.
But those aren't things you can really separate for proprietary models. Keeping inference running also requires staff, not just for the R&D.
This is one of those “don’t be evil” like articles that companies remove when the going gets tough but I guess we should be thankful that things are looking rosy enough for Anthropic at the moment that they would release a blog like this.
The point about filtering signal vs. noise in search engines can’t really be stated enough. At this point using a search engine and the conventional internet in general is an exercise in frustration. It’s simply a user hostile place – infinite cookie banners for sites that shouldn’t collect data at all, auto play advertisements, engagement farming, sites generated by AI to shill and produce a word count. You could argue that AI exacerbates this situation but you also have to agree that it is much more pleasant to ask perplexity, ChatGPT or Claude a question than to put yourself through the torture of conventional search. Introducing ads into this would completely deprive the user of a way of navigating the web in a way that actually respects their dignity.
I also agree in the sense that the current crop of AIs do feel like a space to think as opposed to a place where I am being manipulated, controlled or treated like some sheep in flock to be sheared for cash.
The current crop of LLM-backed chatbots do have a bit of that “old, good internet” flavor. A mostly unspoiled frontier where things are changing rapidly, potential seems unbounded, the people molding the actual tech and discussing it are enthusiasts with a sort of sorcerer’s apprentice vibe. Not sure how long it can persist, since I’ve seen this story before and we all understand the incentive structures at play. Does anyone know how if there are precedents for PBCs or B-Corp type businesses to be held accountable for betraying their stated values? Or is it just window dressing with no legal clout? Can they change to a standard corporation on a whim and ditch the non-shareholder maximization goals?
There’s nothing old internet about these AI companies. Old internet was about giving out and asking for nothing in return. These companies take everything and give back nothing, unless you are willing to pay that is.
I get the sentiment, but if you can't acknowledge that AI is useful and currently a lot better than search for a great many things, then it's hard to have a rational conversation.
why do they need to acknowledge something outside of the point they're trying to make?
Because it was a middlebrow dismissal of the GP
because that's how conversations work. anything less is sparkling debate.
how is it useful to be fed misleading nonsense?
No, they don't. They soak up tons of your most personal and sensitive information like a sponge, and you don't know what's done with it. In the "good old Internet", that did not happen. Also in the good old Internet, it wasn't the masses all dependent on a few central mega-corporations shaping the interaction, but a many-to-many affair, with people and organizations of different sizes running the sites where interaction took place.
Ok, I know I'm describing the past with rosy glasses. After all, the Internet started as a DARPA project. But still, current reality is itself rather dystopic in many ways.
Just enjoy the "good times" powered by other peoples money.
> This is one of those “don’t be evil” like articles that companies remove when the going gets tough but I guess we should be thankful that things are looking rosy enough for Anthropic at the moment that they would release a blog like this.
Exactly this. Show me the incentive, and I'll show you the outcome, but at least I'm glad we're getting a bit more time ad-free.
And it's very timely and intentional, as Gemini is already shoveling product links on my face repeatedly, while OpenAI is testing ads recently. [0]
[0] https://openai.com/index/our-approach-to-advertising-and-exp...
Right, if there's no legal weight to any of their statements then they mean almost nothing. It's a very weak signal and just feels like marketing. All digital goods can and will be made worse over time if it benefits the company.
> Introducing ads into this would completely deprive the user of a way of navigating the web in a way that actually respects their dignity.
Say what you will, there are at least ad blockers for ads on the internet. There are _no_ ad blockers for ads in chatbots.
> I guess we should be thankful that things are looking rosy enough for Anthropic
Forgive me if I am not.
I agree, but at least this is a policy. "Don't be evil" was vague bullshit.
Current LLMs often produce much, much worse results than manually searching.
If you need to search the internet on a topic that is full of unknown unknowns for you, they're a pretty decent way to get a lay of the land, but beyond that, off to Kagi (or Google) you go.
Even worse is that the results are inconsistent. I can ask Gemini five times at what temperature I should take a waterfowl out of the oven, and get five different answers, 10°C apart.
You cannot trust answers from an LLM.
> I can ask Gemini five times at what temperature I should take a waterfowl out of the oven, and get five different answers, 10°C apart.
Are you sure? Both Gemini and ChatGPT gave me consistent answers 3 times in a row, even if the two versions are slightly different.
Their answers are inline with this version:
https://blog.thermoworks.com/duck_roast/
What do you mean, "are you sure"? I literally saw and see it happen in front of my eyes. Just now tested it with slight variations of "ideal temperature waterfowl cooking", "best temperature waterfowl roasting", etc. and all these questions yield different answers, with temperatures ranging from 47c-57c (ignoring the 74c food safety ones).
That's my entire point. Even adding an "is" or "the" can get you way different advice. No human would give you different info when you ask "what's the waterfowl's best cooking temperature" vs "what is waterfowl's best roasting temperature".
Did you point that out to one of them… like “hey bro, I’ve asked y’all this question in multiple threads and get wildly different answers. Why?”
And the answer is probably because there is no such thing as an ideal temperature for waterfowl because the answer is “it depends” and you didn’t give it enough context to better answer your question.
Context is everything. Give it poor prompts, you’ll get poor answers. LLMs are no different than programming a computer or anything else in this domain.
And learning how to give good context is a skill. One we all need to learn.
But that isn't how normal people interact with search engines. Which is the whole argument everyone is saying here, how LLMs are now better 'correct answer generators' than search engine. They're not. My mother directly experienced that. Her food would have come out much better if she completely ignored Gemini and checked a site.
One of the best things LLMs could do (and that no one seems to be doing) is allow it to admit uncertainty. If the average weight of all tokens in a response drops below X, it should just say "I don't know, you should check a different source."
At any rate, if my mother has to figure out some 10 sentence stunted multiform question for the LLM to finally get a good consistent answer, or can just type "best Indian restaurant in Brooklyn" (maybe even with site:restaurant reviews.com"), which experience is superior?
> LLMs are no different than programming a computer or anything else in this domain.
Just feel like reiterating against this: virtually no one programs their search queries or query engineers a 10 sentence search query.
If I made a new, not-AI tool called 'correct answer provider' which provided definitive, incorrect answers to things you'd call it bad software. But because it is AI we're going to blame the user for not second guessing the answers or holding it wrong ie. bad prompting.
I created an account just to point out that this is simply not true. I just tried it! The answers were consistent across all 5 samples with both "Fast" mode and Pro (which I think is really important to mention if you're going to post comments like this - I was thinking maybe it would be inconsistent with the Flash model)
Unfortunately, despite your account creation it remains true that this happened. Just tested it again and got different answers.
It obviously takes discipline, but using something like Perplexity as an aggregator typically gets me better results, because I can click through to the sources.
It's not a perfect solution because you need the discipline/intuition to do that, and not blindly trust the summary.
Did you actually ask the model this question or are you fully strawmanning?
My mother did, for Christmas. It was a goose that ended up being raw in a lot of places.
I then pointed out this same inconsistency to her, and that she shouldn't put stock in what Gemini says. Testing it myself, it would give results between 47c-57c. And sometimes it would just trip out and give the health-approved temperature, which is 74c (!).
Edit: just tested it again and it still happens. But inconsistency isn't a surprise for anyone who actually knows how LLMs work.
> But inconsistency isn't a surprise for anyone who actually knows how LLMs work
Exactly. These people saying they've gotten good results for the same question aren't countering your argument. All they're doing is proving that sometimes it can output good results. But a tool that's randomly right or wrong is not a very useful one. You can't trust any of its output unless you can validate it. And for a lot of the questions people ask of it, if you have to validate it, there was no reason to use the LLM in the first place.
I'm only 25 and grew up in a post "don't be evil" world so maybe I'm jaded, but it always seems strange to me when companies try to make any moral stand about anything. Presumably if there was something a company could do to make more money they'd do it regardless of the ethical standing behind it b/c
1. The people running the company aren't stupid and want to make more money 2. The company is beholden to investors to make more money both practically and legally
IMO this means the only reason anthropic isn't adding ads is b/c they make more money posturing as the "good guys". As soon as that's no longer true not only are the people in charge going to do it, but there's probably some legal standing that would obligate/compel them to do so
A company is just a group of people. Anything a group of people can believe, a company can believe.
A company with investors is a little more suspect, its true. But even massive public companies like Costco still stake out some claims with values beyond the economic and then pursue them. It's not an impossibility.
I also think, despite dropping the motto, that Google did not in fact ever become anything close to evil. I think people are being somewhat unfair (and hyperbolic at best) by claiming otherwise.
Google is a massive population monitoring and propaganda machine.
With AI and an authoritarian government they are an existential threat to humanity.
Does it matter if they are truly moral or truly cynical but think they can capture some sort of "morality premium" by taking a stance?
Yes they are beholden to investors. But so is Musk. The thing is that his investors are also his fans, so a majority of them will approve whatever tactics or schemes he comes up with.
Anthropic may be able to cultivate investors in a similar way, without even trying. After all, why would you invest in a company if you didn't agree with how it was run? Of course there could be a hostile takeover by private equity group who thinks they can squeeze more value out of the tech, but that is harder the larger the company grows.
Edit: also, the morality premium is usually enjoyed more in terms of employees than investors. Google for a long time attracted idealists, by which I mean they were able to pay slightly below market rate for a certain kind of person. If you have a moral compass and you get an equal offer from oai and anthropic right now, which one you choosing?
The key hurdle for AI to leap is establishing trust with users. No one trusts the big players (for good reason) and it is causing serious anxiety among the investors. It seems Claude acknowledges this and is looking to make trust a critical part of their marketing messaging by saying no ads or product placement. The problem is that serving ads is only one facet of trust. There are trust issues around privacy, intellectual property, transparency, training data, security, accuracy, and simply "being evil" that Claude's marketing doesn't acknowledge or address. Trust, on the scale they need, is going to be very hard for any of them to establish, if not impossible.
Impossible. The only way to know what is happening is to have the code run on your own infra.
That still doesn't mean much unless you're doing your own training or getting the weights from a trusted source, and neither of those mean much without knowing something about the data being trained on.
If someone is trying to influence your results, running the inference on your own infrastructure prevents some attack vectors but not some of the more plausible and worrying ones.
I don't think people are concerned about the models' math being biased/tainted (people know of it but that largely doesn't factor into the "security concerns" that people cite.) Typically, it's about how do we know that our data is not going to be seen by a 3rd party. That's what I'm speaking to. Running on your own infra, you can guarantee there are no phone-homes.
What do you mean? Google is roughly the most trusted organization in the world by revealed preference. The 800(?) million ChatGPT users – I have a hard time reading that as a trust problem.
Usage metrics don't reveal preference in all cases. The fact these companies are sketchy/untrustworthy is practically a meme, including among non-tech people. Their services are widely relied upon, but they enjoy very little subjective good will
I always found Anthropic to be trying hard to signal as one of the "good guys".
I wonder how they can get away without showing Ads when ChatGPT has to be doing it. Will the enterprise business be that profitable that Ads are not required?
Maybe OpenAI is going for something different - democratising access to vast majority of the people. Remember that ChatGPT is what people know about and what people use the free version of. Who's to say that making Ads by doing this but also prodiding more access is the wrong choice?
Also, Claude holds nothing against ChatGPT in search. From my previous experiences, ChatGPT is just way better at deep searches through the internet than Claude.
ChatGPT is providing a ridiculous amount of free service to gain/keep traction. Others also have free tiers, but to a much lesser extent. It's similar to Uber selling rides at a loss to win markets. It will get you traction, yes, but the bill has to be paid one day.
Even when you're subscribed, they're providing unreasonable amounts of compute for the price. I am subscribed to both Claude and ChatGPT, and Claude's limits are so tiny compared to ChatGPT's that it often feels like a rip-off.
Clause isn’t trying to compete with OpenAI in the general consumer chat bot space.
None of the ai companies are, they are all looking for those multi billion deals to provide the backplane for services like Copilot and Siri. Consumer chatbots are pure marketing, no company is going to make anything off those $20 per month subs to ai chatbots.
Besides the editorial control -which openai openly flagged to want to remain unbiased- there is a deeper issue with ads-based revenue models in AI: that of margins. If you want ads to cover compute & make margins -looking at roughly $50 ARPU at mature FB/GOOG level- you have two levers: sell more advertisement, or offer dumber models.
This is exactly what chatgpt 5 was about. By tweaking both the model selector (thinking/non-thinking), and using a significantly sparser thinking model (capping max spend per conversation turn), they massively controlled costs, but did so at the expense of intelligence, responsiveness, curiosity, skills, and all the things I've valued in O3. This was the point I dumped openai, and went with claude.
This business model issue is a subtle one, but a key reason why advertisement revenue model is not compatible (or competitive!) with "getting the best mental tools" -margin-maximization selects against businesses optimizing for intelligence.
The vast majority of people don't need smarter models and aren't willing to pay for a subscription. There's an argument to be made that ads on free users will subsidize the power users that demand frontier intelligence - done well this could increase OpenAI's revenue by an order of magnitude.
This is going to be tough to compete against - Anthropic would need to go stratospheric with their (low margin) enterprise revenue.
This will be an amusing post to revisit in the internet archives when or if they do introduce ads in the future but dressed up in a different presentation and naming. Ultimately the investors will come calling.
History is littered with challenger companies chest thumping that they’re never going to do the bad thing, then doing the bad thing like a year later.
"Don't be evil."
> The goals of the advertising business model do not always correspond to providing quality search to users.
- Sergey Brin and Lawrence Page, The Anatomy of a Large-Scale Hypertextual Web Search Engine, 1998
"OpenAI"
[dead]
They are using this to virtue signal - but in reality it's just not compatible with their businesses model.
Anthropic is mainly focusing on B2B/Enterprise and tool use cases, in terms of active users I'd guess Claude is distant last, but in terms of enterprise/paying customers I wouldn't be surprised if they were ahead of the others.
See Github, which doesn't have display advertising.
History shows that software companies with large chunk of their platform being Free to Use mainly survive thanks to Ads.
It goes well beyond free to use models unfortunately.
I believe Perplexity is doing this already, but specifically for looking up products, which is how I use AI sometimes. I am wondering how long before eBay, Amazon etc partner with AI companies to give them more direct API access so they can show suggested products and what not. I like how AI can summarize things for me when looking up products, then I open up the page and confirm for myself.
Won't all the ad revenue come from commerce use cases ... and they seem to be excluding that from this announcement:
> AI will increasingly interact with commerce, and we look forward to supporting this in ways that help our users. We’re particularly interested in the potential of agentic commerce
Why bother with ads when you can just pay an AI platform to prefer products directly? Then every time an agentic decision occurs, the product preference is baked in, no human in the loop. AdTech will be supplanted by BriberyTech.
if llm ads become a real thing, let’s acknowledge that this is exactly what will happen in no uncertain terms.
The only chance of that happening is if Altman somehow feels sufficiently shamed into abandoning the lazy enshittification track to monetization.
I don't think they have an accurate model for what they're doing - they're treating it like just another app or platform, using tools and methods designed around social media and app store analytics. They're not treating it like what it is, which is a completely novel technology with more potential than the industrial revolution for completely reshaping how humans interact with each other and the universe, fundamentally disrupting cognitive labor and access to information.
The total mismatch between what they're doing with it to monetize and what the thing actually means to civilization is the biggest signal yet that Altman might not be the right guy to run things. He's savvy and crafty and extraordinarily good at the palace intrigue and corporate maneuvering, but if AdTech is where they landed, it doesn't seem like he's got the right mental map for AI, for all he talks a good game.
There are a number of different llms - no reason they all need to do things the same. If you are replacing web search then ads are probably how you earn money. However if you are replacing the work people do for a company it makes more sense to charge for the work. I'm not sure if their current token charges are the right one, but it seems like a better track.
yeah it’s either that or openai has effected a massive own-goal… im leaning toward your view, but hoping that prediction does not manifest. i would be fine with all sorts of shit in life being more expensive but ad-free… but this is certainly a priviledged take and i recognize that.
My thoughts exactly. They are using the Google playbook of "don't be evil" until it becomes extremely profitable to be evil.
[dead]
You really think the giant ad company would put ads into their product after saying they won't? You should strive to be less cynical.
> Anthropic is focused on businesses, developers, and helping our users flourish. Our business model is straightforward: we generate revenue through enterprise contracts and paid subscriptions, and we reinvest that revenue into improving Claude for our users. This is a choice with tradeoffs, and we respect that other AI companies might reasonably reach different conclusions.
Very diplomatic of them to say "we respect that other AI companies might reasonably reach different conclusions" while also taking a dig at OpenAI on their youtube channel
https://www.youtube.com/watch?v=kQRu7DdTTVA
I think this says a lot about the business approach of Anthopic compared to OpenAI. Just the vast amount of free messages you get from OpenAI is crazy that turning a profit with that seems impossible. Anthropic is growing more slowly but it seems like they are not running a crazy deficit. They do not need to put ads or porn in their chatbot
Absolutely! (:
Ads are coming to AI. But not to Claude. Recent advertising campaigns from Anthropic.
Violation https://www.youtube.com/watch?v=kQRu7DdTTVA
Betrayal https://www.youtube.com/watch?v=FBSam25u8O4
Deception https://www.youtube.com/watch?v=De-_wQpKw0s
Treachery https://www.youtube.com/watch?v=3sVD3aG_azw
"Think Different" also used to mean something, before Tim Cook cozied up to authoritarianism. Who knows, it's never too late for Anthropic to have a change of heart.
One can only hope.
They are not trying to sell adds. They are trying to sell themselves as a monthly service. That is what I think when they are trying to convince me to go there to think. I rather go think at Wikipedia.
Idk, brainstorming and ideating is my main use case for AI
I use it as codegen too but I easily have 20x more brainstorming conversations than code projects
Most non-tech people I talk to are finding value with it with traditional things. The main one I've seen flourish is travel planning. Like, booking became super easy but full itinerary planning for a trip (hotels, restaurants, day trips/activities, etc) has been largely a manual thing that I see a lot of non-tech people using llms for. It's very good for open ended plans too, which the travel sites have been horrible at. For instance, "I want to plan a trip to somewhere warm and beachy I don't care about the dates or exactly where" maybe I care about the budget up front but most things I'm flexible on - those kinds of things work well as a conversation.
Wikipedia is, of course very useful, but what it’s not good at is surfacing information I am unfamiliar with. Part of this problem is that Wikipedia editors are more similar to me, and more interested in similar things to me, than the average person writing text that appears online. Part of the problem is that the design of Wikipedia does not make it easy to stumble upon unexpected information; most links are to adjacent topics given they have to be relevant to the current article. But regardless, I’m much more likely to come across a novel concept when chatting with Claude, compared to browsing Wikipedia.
It’s so hard to succeed without selling ads. There’s an exponential growth aspect to these endeavors and ads add a lot of revenue, which investors like, so those who don’t can find that the lost revenue “multiplies” due to lower outside investment, lower stock price growth, etc.
I wish the financial aspects were different, because Anthropic is absolutely correct about ads being antithetical to a good user experience.
Anthropic is very big (the biggest AI co?) in B2B, where you don't have ads. Also, if they end up creating a datacenter full of geniuses, ads won't make sense either.
B2B will be hard to compete with vs Google and MSFT, as they can bundle services with Office365 or Google Workspace.
Sometimes posts like this are just value-signaling. I hear a lot of cynicism and "just you wait, the other shoe will drop" comments along those lines.
But combined with the other projects Anthropic has pursued (e.g. around understanding bias and explaining "how the model is thinking as it is") and decisions it has made, I'm happy with the course they're plotting. They seem consistently upstanding, thoughtful, and respectful. I want to commend them and earnestly say: Keep up the good work!
They made an ad to say that they won't have ads, i dont know if they are aware of the irony.
https://x.com/ns123abc/status/2019074628191142065
In any case, they draw undue attention to openAI rather than themselves. Not good advertising
Both openAI and Anthropic should start selling compute devices instead. There is nothing stoping open-source LLMs from eating their lunch mid-term
Ads as a concept are not evil. There have been ads since prehistory.
Littering a potentially quality product with ads which one cannot easily separate is what the evil is.
What an awful take
Why? Give me the awesome take
Non factual or interesting. No - ask an LLM I'm sure you'll be interested
Why is it ironic?
It would seem to be ironic if they put ads in their models to say they won't put ads.
> "I'm very interested in the idea of a piece of music being a place to think," Richter explained, adding that he had written Daylight as a response to the 2003 Iraq War. [0]
0. https://www.npr.org/2020/01/22/796801746/max-richter-tiny-de...
Seems short-sighted to commit to not running ads exactly like OpenAI plans to. You win in the court of public opinion for a few months, then look like a hypocrite when you're inevitably forced to run ads because your investors demand it. Sort of like how safetyism was convenient marketing until it became clear that it was revenue repellant and they quietly walked it away.
I asked for this last week in an hn comment and people were pretty negative about it in the replies.
But I’m happy with position and will cancel my ChatGPT and push my family towards Claude for most things. This taste effect is what I think pushes apple devices into households. Power users making endorsements.
And I think that excess margin is enough to get past lowered ad revenue opportunity.
So apparently they're going to run a Super Bowl ad about ChatGPT having ads (without saying ChatGPT of course)........ Has doing an ad that focuses only on something about your competitor ever been the best play? Talk about yourself.
Obviously it's a play, honing in on privacy/anti-ad concerns, like a Mozilla type angle, but really it's a huge ad buy just to slag off the competitors. Worth the expense just to drive that narrative?
Ads playlist https://www.youtube.com/playlist?list=PLf2m23nhTg1OW258b3XBi...
Wasn't Apple's iconic 1984 ad basically that?
Apple's ad had a woman dressed like a Hooter's waitress to represent themselves. That makes themselves the focus of attention.
https://www.youtube.com/watch?v=ErwS24cBZPc
That's a pretty weird take. Her costume is typical of track and field athletes of the time, except that her jersey represents the nation of the Apple Macintosh, which you can barely notice in the ad because it flashes by in a split second.
ah, good one. Was it Big Blue or Big Brother in general being referenced in that one? Either way I suppose Apple didn't even say much of anything about their product in that one where Anthropic is at least highlighting a feature.
People understood blue Big Brother represented Big Blue. And Steve Jobs named IBM when he discussed the ad.
> There are many good places for advertising. A conversation with Claude is not one of them.
> ...but including ads in conversations with Claude would be incompatible with what we want Claude to be: a genuinely helpful assistant for work and for deep thinking.
Sadly, with my disillusionment with the tech industry, plus the trend of the past 20 years, this smacks of Larry Page's early statements about how bad advertising could distort search results and Google would never do that. Unsurprisingly, I am not able to find the exact quote with Google.
Yeah, it’s a shame we’ve all grown so jarred, I do see this better than nothing.
In this animal farm Orwellian cycle we’ve been going through, at least they start here, unlike others.
I for one commend this, but stay vigilant.
I appreciate taking a stance, even if nobody is asking. It would be great if it was less of a bad faith effort.
It's great that Anthropic is targeting the businesses of the world. It's a little insincere to than declare "no ads", as if that decision would obviously be the same if the bulk of their (not paying) users.
There are, as far as ads go, perfectly fine opportunities to do them in a limited way for limited things within chatbots. I don't know who they think they are helping by highlighting how to do it poorly.
Not the point of the article, but just responding to the headline: LLMs are a space to think in the same way that watching television is a space to think. Some people may think while watching TV, maybe many of the people reading this watch TV this way, but most will not. In fact, television mostly acts as a soporific, killing intelligent thought and numbing the mind against any sort of complex cognition. Same with LLMs. They're a space to sleepily watch someone else think.
I make them ask me questions. Forces me to think.
So they have "made a choice" to keep Claude ad-free, they say. "Today [...] Claude’s only incentive is to give a helpful answer", they say. But there's nothing that suggests that they can't make a different choice tomorrow, or whenever it suits them. It's not profitable to betray your trust too early.
I can't really imagine any statement they could give that would ease concerns that at some point in time they change their mind. But for now, it is a relief to read, even if this is a bit of marketing. The longer it goes without being enshittified the better.
They could agree to some actual significant negative consequence to running ads. e.g. They could put a clause in the subscription signup process that says if they ever run ads - even if it's only for free accounts - then you get all of the money you've spent back.
Of course I realise they would never do something like that. Buy why not? Well, because they might decide they want to run ads...
Sure, ad free forever, until it is not.
Great by Anthropic, but I put basically no long term trust in statements like this.
Ben Thompson has long been insistent that ChatGPT and other AI tools basically have to have ads and it's been a big mistake they didn't have them sooner. It's an interesting take:
> What I think is clear is they have to build an advertising product, and the reason they have to build an advertising product is any consumer Internet product has to be advertising, because it’s such a beneficial model to everyone involved, and the reason it’s so beneficial is you get to indefinitely and infinitely increase average revenue per user without any worries about price elasticity, because the entire increase in average revenue per user is borne by the advertisers who are paying it willingly because they’re getting a positive return on their investment, and everyone’s using it for free so you can reach the whole world. Then what happens with that is once you get that model going, you have a massive R&D advantage, because you have so much more money coming in than anyone who doesn’t have that cycle or who has to charge users for it.
https://stratechery.com/2026/ads-in-chatgpt-why-openai-needs...
> This point, more than anything else, explains why the company so desperately needs an advertising model. Advertising is the only potential business model that can meaningfully bend the revenue curve such that the company can not just fund its compute but gain leverage on it, for all of the reasons I laid out before: first, advertising increases the breadth of the business, in that you can offer a better product to more people, increasing usage and expanding inventory. Second, advertising increases the depth of the business, in that there is infinite upside in terms of average revenue per user: more usage means more inventory on one hand, and building out the capability for effective targeting and high conversion rates increases the amount that advertisers are willing to pay — even as the cost to the user remains the same (ideally free).
It's valuable to remember that advertisers will pay more per user than users will, and that's hard to beat in a competitive market.
100%. Love this approach by Anthropic. The Meta "monetization league" is assembling at OpenAI and doing what they've done best at Meta.
However, I do think we need to take Anthropic's word with a grain of salt, too. To say they're fully working in the user's interest has yet to be proven. This trust would require a lot of effort to be earned. Once the companies intends to or becomes public, incentives change, investors expect money and throwing your users under the bus is a tried and tested way of increasing shareholder value.
I have such mixed feelings about Anthropic. Opus 4.5 is legit, Claude Code has been making solid progress (but it's gonna get stomped by pi) and they've been trying to be more transparent. On the other hand, they still infra issues, they till haven't open sourced Claude Code and they're going after other harnesses that use their sub. Dario also seems to have a very paternalistic attitude that I don't like, though I don't think that will survive the market.
I really want to applaud Anthropic; I remain cautiously optimistic, but I’m not certain how long they will maintain this posture. I will say that the recent announcement from OpenAI has put me off from ChatGPT — I use Gemini occasionally, because it’s the devil I know. OpenAI has gone back and forth on their positions so many times in a way that feels truly hostile to their users.
Plus, I’m not a huge fan of Sam Altman.
What makes Anthropic seem like early Apple is not just the unique taste, but the courage to stand firm with their vision of what the product should be.
That courage was nowhere to be found when Palantir rolled up with a truckload of cash.
What's the problem with Palantir?
I am going to assume that is a question asked in good faith by an inquisitive mind. That assumed, asking that question to any proper LLM will provide an answer more in depth than I am willing to type this evening.
Palantir, to me, is the weaponization of big data, where advanced analytics are used to target vulnerable populations. Not just abroad, but here against its own citizens. It is the dystopic enabler that we have been warned about.
Palantir and the words from its leadership seem to me to be in direct opposition to parts of the Constitution doc that Anthropic hold up to show their ethics and seriousness.
It’s better to not fall for serif fonts and warm colors.
Apple had a vision, all right. It was our fault that we thought they would become the rebel with the hammer, and not the guy on the screen.
Only 4 years old, they haven't existed long enough to be "firm".
Making formal, public statements like this is a good start. It is certainly better than NOT making these sorts of statements.
Yeah. Does anyone remember how long did it take GOOG to remove "Don't be evil" from their motto?
> Anthropic seem like early Apple
sorry but this is silly, nothing suggests this at all.
>An advertising-based business model would introduce incentives that could work against this principle.
I agree with this - I'm not so much worried that ChatGPT is going to silently insert advertising copy into model answers. I'm worried that advertising alongside answers creates bad incentives that then drive future model development. We saw Google Search go down this path.
No ads in AI? cheap commitment
Ads are more then obsolete
Using ads in AI, today, is like printing flyers when the Internet started, (or sending email ads 30 years ago: of course Google could have promised: "we will not send spam to your gmail mailbox..." :).
"Ads" aim to influence your behavior: AI is a doing much much more with no need for ads (Claude included)
Claude have posted on number of very sarcastic videos on twitter that take a jibe at ads https://x.com/claudeai/status/2019071118036942999 with an ending line "Ads are coming to IA. But not to Claude."
I do feel like many folks here and elsewhere are missing the mark with LLMs.
1) Yes, they are absolutely useless in a consumer setting. 2) If you want to be a software developer, you absolutely need to know how to understand/interact with one, and you more than likely will need to understand things like https://continue.dev.
I am no longer in software development due to my body slowly (quickly) dying, however I see it all from the sidelines:
1) New tech was rushed to the front lines way too quickly by big tech. 2) Big (and small tech) rushed layoffs way too fast rather than let we devs explore the advantages vs. disadvantages. 3) Companies blame "AI" (LLMs) for layoffs. 4) Most senior devs (including myself) soundly reject AI due to the above. 5) New generation of devs uses AI tools, some struggle occurs where morons don't bother reviewing code that was written by an auto completion engine. 6) We nerds begin to understand the usefulness of LLMs for "the boring part"
Not a shareholder of any company. I'm permanently disabled. Just watching this stuff from the sidelines.
What other interaction models exist for Claude given that Anthropic seems to be stressing so much that this is for "conversations"?
(Props for them for doing this, don't know how this is long-term sustainable for them though ... especially given they want to IPO and there will be huge revenue/margin pressures)
Does the veneer of goodness despite (alleged) cutthroat business practices from Anthropic bother anyone else?
It's one of those marketing stunts where they take a dig at another company, then a year or so later find themselves doing the same thing.
Think of the Samsung headphone jack commercial.
Anthropic probably saw how much money they made off of the Moltbot hype and figured that they don’t need ad revenue. They can go a step further and build a marketplace for similar setups, paying the developers who make them in micro transactions per tokens.
I actually do want ads. Subtle, labeled ads. For many queries in Google, I like to know who is advertising for an intent. In may cases, that is a perfect fit. I don’t want ONLY the companies who have gamed the LLM SEO market to win.
Good on Anthropic! I appreciate how deliberate they are on maintaining user trust. Have preferred Claude's responses more through the API, so I don't imagine this would have affected me as much but it is still nice to see.
Don't understand why more companies don't just make ads opt-in as a trade for more features
A lot of people are ok with ad supported free tiers
(Also is it possible to do ads in a privacy respecting way or do people just object to ads across the board?)
I would object to ads across the board in this case (though I’m generally fine with even targeted ads). It would create a customer-client relationship between companies paying to advertise and the AI company, creating an incentive for Anthropic to manipulate the Claude service on their behalf. As an end user that seeks input from Claude on purchasing decisions, I do not want there to be any question as to whether or not it was subtly manipulated.
It's nice that they don't show ads in conversations with Claude - but I wonder if they collect profiling information from my prompts and activities to sell to advertising firms.
I like their models. They're powerful. But...
Who do they think believe the whole "don't be evil" in 2026?
We know what's around the corner. Enshitification, loss of trust, frog boiling, account restrictions and upsell, advertising, degradation of service, data sold for advertising and worse.
Y'ain't kidding anyone with this stuff. You're only providing screenshots for future memes.
That's true. CI in all of my conversations with AIThat's true. In all my conversations with AI, I think CIaude's thinking is the richest.
They'll shout this message right until the day they change course. Then who will hold them to account?
i spend most of my time with claude thinking about when my daily usage limit is going to reset
Claude focuses on enterprise and B2B rather than mass consumer, so it makes sense for them.
Important to note Anthropic has next to no consumer usage
Wrong (in trumps voice)
From Sama "More Texans use ChatGPT for free than total people use Claude in the US, so we have a differently-shaped problem than they do"
Facts don't care about your feelings
Sama lies all the time.
That’s positive. How is Claude? Is it censorship heavy?
If you broach subjects Anthropic considers sensitive (cyber security, dangerous biotech, etc) Claude is very likely to shut you down completely and refuse to answer. As someone that works in cybersecurity and uses Claude daily, it is annoying to ask a question regarding some feature of Cobalt Strike and have it refuse to answer, even though the tool’s documentation is public. I would have cancelled my ChatGPT subscription at this point if once or twice a month I didn’t need to ask it to look up something when Claude refuses.
How are the Chinese models in this regard? Qwen3 for instance?
Claude is the last place where thinking happens.
Get back to us when you’re profitable.
You need a mind mapping tool there
Since when does HN welcome blatant self-advertising posts like this one ?
RemindMe! 2 years
Let's savour this time pre-enshittification of AI. One day we will look back so fondly at it and wonder at our naivety.
[dead]
> its not about getting answers, its about having a patient collaborator
Looks like you're picking up LLM speak too!
https://www.theverge.com/openai/686748/chatgpt-linguistic-im...
You mean human speech that LLMs were modeled after? Or was JFK influenced by LLMs, too?
There are certainly more tells and more people are speaking like LLMs now where they hadn't before.
I agree that there are a ton of tells, but I wouldn't say picking up some new vocabulary is "speaking like" just yet. It's still early, though, so more studies are needed.
!remind me in two years
/s
[flagged]
[flagged]