In some cases, workers are also being asked to automate the parts of their jobs they enjoy most, Hinds said on the podcast, pointing to customer-service employees who enjoy building relationships but are increasingly expected to supervise AI agents instead.
"That's what gives you joy and meaning at work," she said. "That is very dangerous."
What's a 20% productivity gain if I constantly feel deflated by work that used to energize me? That's going to give back the productivity gain and more, while also decreasing my quality of life.
This is an important point. My light-bulb moment was when I talked to a product owner in a previous job, and I expressed surprise around an expensive planned change, because it didn't seem that valuable to our customers.
He said, "Almost half of what we do is not that valuable to our customers, but it's valuable to him, and her, and him", pointing through the conference-room window at my fellow programmers, "and that's why we do it. If we only did things that were very valuable to our customers, we wouldn't have nearly as many good engineers on the team as we do."
Do the engineers not derive enjoyment in their jobs from making the customer experience better? This does track with what I've experience in my career, where we've gone from everything being to better the user experience to tech companies sort of trying to out-do each other in their technical solutions while the software continuously gets worse and more antagonistic.
Plenty of value comes from things the customer doesn't care about.
Customers want features as fast and as cheap as possible. I derive joy from solid test suites that avoid me getting paged while on call and team processes that don't allow config changes on Friday so pages don't happen on the weekend.
Very few craftspeople derive their joy from the customer experience. An electrician isn't happy because their work allows me to watch TV. A carpenter isn't happy because a new set of stairs lets me get to the basement faster. They're happy because of their perception of the quality of their work. This goes away when the visible or fun parts are no longer "their work"
I wouldn't agree with that. The issue with software is that the people you make things for are usually anonymous and you'll never meet them, but if you've ever built software that helped someone and you witnessed it, it feels really good.
Meeting your customers feels good. It's still not why I do it. I've written components that were used by billions of people and generally loved (pieces of the Windows 7 taskbar) and talked to people about that work. I've worked in the education space and talked directly to the staff at schools who use my product to hear their use cases, thoughts, and feedback. It's fun, humbling, and rewarding. It's not what motivates me day to day.
I don't agree with your statement. Individual productivity is very different from team productivity. AI is much like the traditional cowboy coder - prioritizing speed above all else leads to disaster.
As to customers, customers value stability. They don't want the UI changing every week, they don't want random changes of the data model that corrupt their content, and so the traditions of quality, strong testing and integration pipelines, and so on are just as important in established products - in some ways more important than ever, as there are plenty of opportunities for your competitors to burn user trust and drive them away.
I have yet to encounter an established product demonstrating a massive improvement in velocity due to AI. Being able to spit out prototypes doesn't matter all that much.
I'm handling the AI wave by holding on to my sanity and skills and working to establish a reputation in preparation for big paychecks fixing the messes everyone else is creating with their tokens.
This is the danger of isolating engineering from customers, or even internal customer-interfacing employees.
If all they see is code, they will get satisfaction from tidy code, not user happiness. One good thing about AI is it elevates product engineers because they more directly bridge the customer-product-code divide.
That's my take as well... the dev effectively takes on QA/QC and PM roles as a team working with AI for the baseline of development work. Of course, this is also a slightly different skill set and cognitive load. It also needs to completely upend how project planning happens when you are using Dev+AI in coordination.
Not just isolating them from customers but also from engineering-adjacent work that isn't code.
I've been at a place that is basically microservices slop (several dozen services per engineer). They're all poorly maintained and at least a solid 40% of all this code that they've written could have been just a traefik or nginx configuration/container.
When you have a lot of inexperienced (relative to industry) and overworked software engineers, the solution to every problem becomes to write code and writing new code should be a last resort.
Worse still, there's just a poor general understanding of the internet protocols they're working with and of how to do distributed systems right. Unfortunately with LLMs I've been seeing this get worse, not better.
They use the LLMs for code generation but not architecture review. Bad ideas are getting fully-baked quickly before anyone with good sense can intervene.
There are plenty of projects that are green lit that have good intent but are bone headed when it comes to solution and implementation. Good engineers hate these types of projects. Good PMs try to avoid these at all costs but sometimes your hand gets forced because some VIP, either internal or external volun-tells you to do it.
> Do the engineers not derive enjoyment in their jobs from making the customer experience better?
Quite a few don't, no.
Different people derive enjoyment in different things and some of the best engineers do not find satisfaction in "delivering better customer experience" but in working with, and improving, cool technology. Its up to management to find areas of the business where they can deploy these people in a way that dove-tails with business success.
Its also the case that only working on projects that "deliver customer value", and having to justify every single endeavor through that lense, is how you end up in a local maxima in your tech stack, get mired in technical debt, and then get lapped by your competitors who have the foresight to work on foundational technology that enables future velocity.
To be frank, its endlessly frustrating that your median Hacker News poster doesn't get this, and instead prefer to brow-beat people about how they're caring about the wrong things.
Platform and internal dev teams have customers as well. I'm not terribly frustrated that you don't get this. I've certainly worked with devs (and managers) who wanted to push new technology for the sake of using new technology, and they should have found a side project as an outlet for this.
The person you're responding to never said those teams didn't have customers.
It's not about new technology for the sake of new technology, it's about taking pride in one's work and what that person created.
Honestly, the American obsession with "everyone should think of what the customer wants" is exhausting verging on toxic. The people who talk about that point loudest are inevitably owners saying "you should all care more to make me richer". If you want your employees to care about the customer more than their own personal satisfaction, give them significant equity and significant autonomy such that they can see how their actions have direct impact. Saying they should think of the customer and then treating the employees as an interchangeable cost to be minimized is insulting and won't lead to anyone focusing on the customer.
> The person you're responding to never said those teams didn't have customers.
> It's not about new technology for the sake of new technology, it's about taking pride in one's work what what that person created.
Thank god, I really want to say that I appreciate that you got what I said. A simple upvote didn't feel like enough.
Employees are humans, not robots. Its inconvenient, but if you want a world-class team then you're going to have to deal with the fact that people derive satisfaction from different things, and you're not going to be able to motivate them by beating them into submission about what they "should care about." This may involve having to think creatively about how to manage your people instead of treating them as fungible work units.
I'm going to apologize in advance for being long-winded, but I feel there's a lot to unpack here.
> Platform and internal dev teams have customers as well. I'm not terribly frustrated that you don't get this.
Respectfully, this response here is a perfect example of you not getting it and assuming that I am the one that doesn't get it. You have not said anything that I have not already heard and understood before. The fact that people have different values does not mean they "don't get it." But saying something like "Do the engineers not derive enjoyment in their jobs from making the customer experience better?" does imply that you don't understand other peoples' values.
The fact that you posted "Platform and internal dev teams have customers as well." indicates to me that you missed the point. Whether they are on those teams and whether they consider other engineers their customer is besides the point; they may not derive satisfaction from "delivering value" to those people regardless. That doesn't mean they don't care about their customers, which is the take away the median HN poster takes, but rather that they are not energized and motivated at the end of the day by delivering value to them.
> I've certainly worked with devs (and managers) who wanted to push new technology for the sake of using new technology,
Sure, everyone has. But the flip-side of this is a class of people who assume any tech improvement that doesn't directly move a metric is just an effort at resume-building. Just as often I've seen efforts and building a more robust system as unneeded resume building despite clear need (usually because the need is very hard to measure).
> and they should have found a side project as an outlet for this.
I mean, this is incredibly dismissive and exactly the attitude I was talking about. No-one is saying that engineers should be allowed to just do whatever to have fun. Work is work. But ideally you find ways to organize your team so that everyone is motivated and energized by their work, and doing so requires that you understand that not everyone is motivated by the same thing. But in these discussions, the attitude comes across as "everyone should be motivated by delivering good customer experience and if they aren't we shouldn't care."
If there's no opportunity to give these sorts of people fulfilling work, then fair enough. It *is* work. But the attitude displayed here is that we shouldn't even try and understand their values and think about ways to productively deploy that.
As an aside about customers, internal and external customers are, in my experience, treated vastly differently. We care about experience for external customers, but internal customers is usually all about velocity and trade offs. The bar is substantially lower, and rough edges are almost always ignored. So I am skeptical at the idea that we can just frame internal users as customers and all the discrepancies go away.
It also misses the fact that other people on my team are also my customers, because they have to maintain the system! And I am also my own customer, because I also have to maintain it!
Force me to click hundreds of buttons per release and I'm going to be disinclined to go through that. You wouldn't have a surgeon have to go hunting for the right tools.
I dunno about everyone, but I do. It drives me up the wall when software is slow, stutters, etc. There’s almost no good excuse for it. But it happens anyway because of some combination of, “nobody cares,” and, “money.”
I derive joy out of people using a thing i helped design and they never get frustrated with it. It works, it’s fast, it solves the problem and it’s rarely a pain to maintain or extend for me or my team. Thats good, solid stuff right there.
That's kinda software development ... generally and not really new as far as I'm concerned.
Even everyone earnestly trying to identify the feature that IS IMPORTANT to clients and will actually be used and appreciated frequently is something of a crap shoot.
A better way to put it is that these things do have value to the customer, the customer just doesn't have a way to understand how the work you're doing provides value because it's the part of the product you don't see. If you clean up technical debt, improve test coverage, improve your deployment systems, etc, it doesn't change the immediate customer experience in a meaningful way, but it does allow you to deliver the changes that customers do see faster and with fewer risks.
This quote makes it seem like the work is self-indulgent, and I have seen that happen sometimes, but it's not half of what we do.
I always found pair programming a bit.. hellish. Chatting about things. Rubberducking. Playing code review ping pong. All great. But the feeling of being shoulder surfed killed my ability to play around with the freedom that I do when I'm alone -- and that playing around sometimes led to better/more interesting outcomes than I'd have gotten otherwise.
Yeah, I always bought the premise that it was good for the software, but it didn't work for me as a person, it drained my energy way too fast to spend hours of the day having to be "on" in conversation with another person.
In many ways, I think what's working for me with AI is that it is very similar to pair programming, but without the social-emotional investment required to interact with another person for long periods of time.
How this goes is the engineers start raising arguments in meetings full of nerdy technical terms such as “refactor”, “technical debt” and “accidental complexity”. As time progresses, more of them say more of this. At some point management learns that when engineers say engineery words like this you sometimes gotta say yes and let them do it even if it means urgent features are delayed, or the engineers will walk.
Good engineers would communicate effectively of something is impacting the product. Good product understands how to judge that impact and prioritize it. Frankly, if the engineers are making enough tech debt that they don't want to play in their own sandbox as part of regular practice, maybe they should walk.
Let them walk. Good programmers can write code that doesn't degrade the customer experience.
In the last 15 years it's been all about the programmer's experience and user-facing software is objectively slower, buggier, and more resource intensive.
It's a two ways street though, isn't it?
With lazy/unskilled/unmotivated developers you will have a shitty product, or maybe your SaaS would explode the day it's starting to get traction.
When I was given a semi-ultimatum "use AI or get fired" kind of thing for writing code I had a brief bout of depression/sadness. Whereas my friend doesn't care/says "I get paid to not work". I have gotten past it, now I'm just like, I'll do what I need to do to get paid since unfortunately I'm in a lot of debt so I need this job. I learned to code in 2013 so I like typing the code myself but now it seems like a waste of time. I still write my own code for myself/hardware hobby.
If you're anything like me, and you stay in a coding/dev/IC track for your career because you like the work, you will eventually hit a point where you start thinking it's all meaningless. This happened before AI for me, but AI certainly reinforces it.
You come to a point where you realize that you're not doing anything that creative, or nothing you haven't done hundreds of times before, maybe every few years you switch to whatever new tech stack has gotten popular, but it's fundamentally all the same. And you start to realize that everything you do has a lifespan of a few years, and then you (or probably someone else) will re-do it.
As retirement starts feeling like it is something that will happen sooner than later, you look back and see that almost nothing you've built is still in use, or will be for very long after you're gone.
I hope to retire in about two years. At that point, I plan to not be using any technology or computers in my life for a while, or as little as possible. Maybe at some point I'll rediscover some of the fun I used to have writing programs for myself, but I suspect I'll need a long break before that happens.
> maybe every few years you switch to whatever new tech stack has gotten popular, but it's fundamentally all the same
So true!
But it's interesting that, from the perspective of someone in the middle, neither near the beginning or end of my career, I am (now, after a period of sadness) experiencing AI as a reinvigoration of fun in the work. But it's a very different kind of fun. I had totally lost the fun of clean code and figuring out new technologies and approaches and abstractions, just like you describe.
But now I'm experiencing the joy of thinking about what I can build, now that it's so much faster and easier to try ideas. I think this is actually getting back to an earlier version of my joy with computers. I can (vaguely) remember in my early years being like "wow! cool! I can make stuff that shows up on a computer screen!". But then it turned out to be ... pretty damn hard to actually do that, which led me to more excitement about all the ideas and technologies and techniques for managing the complexity of software engineering. But then that started feeling more tedious and samey, but I still had to put lots of time into it, there wasn't any other option.
But now all that is so much easier, and I'm rediscovering the fun of "wow cool, I can make things!", but now also with the whole benefit of the time I have spent doing the work of software engineering.
> you will eventually hit a point where you start thinking it's all meaningless ... You come to a point where you realize that you're not doing anything that creative
I hit that point after about 5 years. At the time, software development was mostly desktop applications (which are sadly defunct), and UI frequently doesn't require much in the way of algorithms. Then I focused on the challenge of design a UI that matches the user's mental model, which lasted about five years. After another a stint in mobile games and apps, I started doing contracting. The key for me was the concept of serving someone else' vision (instead of my own vision of FIRE), and some HN conversation on some of Steve Job's videos on work. This all boiled down into an attitude of "I will solve your problem (and you will pay me)" (from Job's discussion with the guy who designed the NeXT logo).
As a contractor, sometimes I solve your problem with a cool algorithm, sometimes by simply implementing your pixel-perfect design, and sometimes with a true-Agile back-and-forth of vague design into concrete software. Since I switching to the "I solve your problem" perspective, I haven't had any problems with feeling frustrated at just being a code monkey implementing stuff I don't need to think about. (However, I have discovered that I get really bored doing server stuff, and it doesn't save me from that. I just like something concrete, and not having to debug by putting in print statements and wading through thousands of lines of output, with only the reward that the variable contents is now correct. Turns out I want to see something. I get unreasonably excited about going from a blank screen and slowly populating it with widgets.)
I've definitely found what you're describing at bigger companies, but I also previously had experience writing software at smaller non-technology companies.
Legal marketing specifically. Weirdly, my work had more impact, respect and longevity there than the place where I'm a much more senior engineer supposedly directing the work of a whole organization of engineers. I had it better where I was a 1 of 2 than a leader among hundreds.
Small successful companies are great, but the hardest thing for me psychologically has been when I'm at a small company that is struggling to convince anyone to use its product. Being a small cog in a giant machine serving lots of users is more satisfying (to me) than building things that nobody is using.
I was exactly this way ... maybe a year or so ago.
But honestly I've stopped being excited to type out code in my personal projects anymore either. I've become much more excited by what I can accomplish on my own in a small number of hours squeezed between work and family. I still experience this as a loss, but I'm no longer so sad about it, and moreso feel invigorated by the possibilities and opportunities that have been opened up.
A way that I have come to think about this is: I used to always be curious about the product management role. How exciting to come up with ideas and validate them with users. But I always demurred because it would be so frustrating to have to rely on other people to bring those ideas to fruition! On balance, I always preferred being the one executing ideas to being the one generating and validating them. But now I can properly do both things! (In my hobby time, that is, at work we still have this idea/execution split, at least for the time being.)
FWIW, I was just like you but then completely gave in and found enjoyment in the act of simply ideating and shipping. The gap now between idea and implementation is so small. At first I was depressed but now I'm in the acceptance phase of grief. We aren't going back, for better or for worse.
Heh, my employer kept pushing us to use Copilot. And over the last months the cli has actually gotten halfway decent... So I did start using it. Albeit sparingly because the token allotment was always pretty low.
Then they announced that they removed the limit/making further request just cost extra for them. That's when I started using it as I did for my personal projects I pay subscriptions for...
Then Copilot increased their pricing. Announced in April I think? But took effect this month. This Monday they announced that the limits are back in effect. So I guess I'll be going back to hand coding next week, as my tokens are about to run out ಥ ‿ ಥ
Corporate is always so silly. I mean I know how it happens: everyone just wants to get their bonus, so different management roles try to coerce the employees to do whatever best serves their bottomline - rarely related to whatever is good for the corporation... But it's always silly to live through it.
I think this is a larger question we should be asking. YES, we can build this world: a world where robots do our chores, serve our coffee, check us out at the grocery store, and let the AI agents do the parts of our jobs we love.
But SHOULD we? With great power comes great responsibility - and I'm getting the impression we're (quickly) building a world that isn't very fun to live in. We technically have a choice here - DO we want bots writing our prose and responding to our customer service inquiries?
I remember a book I read as a pre-teen, 40 or so years ago, about a kid who wanted to be "perfect". A wear a tree of broccoli on a string around your neck to learn how to overcome embarrassment. A perfect person never makes mistakes, and the best way to not make mistakes is to not do anything. Similar "requirements" of perfection and their expression are presented. The kid eventually finds himself in an empty room, by himself, doing nothing, wearing broccoli. Perfection was achieved, but at the cost of an extremely boring life.
I'm of the opinion that "true art"/cultural artifacts can't be automated by definition, as they derive their value from the human experience embedded in them by their creators.
OTOH, I think we absolutely SHOULD automate necessary "drudgery" type work wherever we can, but we're going to need a radical reconceptualization of how we distribute the spoils of economic productivity as a result.
Unfortunately, I think the type of reconceptualization we'd need would entail a complete overhaul of many long-established and deeply-internalized concepts (rights and duties of ownership of intellectual and private property, decoupling of identity and occupation, etc.), and from everything I've seen, that will be a long and painful process assuming it's even achievable. (Especially in the US, where decades of pro-business messaging has yielded a culture that equates income-earning ability/entrepreneurial success with individual human worth. I really struggle to imagine a path toward unwinding that, but there's little chance it'll be a smooth ride.)
> customer-service employees who enjoy building relationships but are increasingly expected to supervise AI agents instead.
It sucks for the employees, otoh it might be the only way we're going to beat Baumol's Cost Disease.
In the past few decades productivity has exploded, but service employees have largely failed to increase productivity in any way because it's harder to automate these tasks.
It's the reason the costs of things like education and healthcare are downright extortionate, the reason you're paying back your college well into your fifties, the reason you don't call an ambulance for someone in the US because you don't want to ruin their life financially.
We may have to trade the personal fulfillment in these jobs for the broader affordable access to these services.
Education and healthcare are both ridiculously overpriced in the US for reasons that have little to do with service costs. Questionable financial systems behind these services are much more to blame.
>It's the reason the costs of things like education and healthcare are downright extortionate, the reason you're paying back your college well into your fifties, the reason you don't call an ambulance for someone in the US because you don't want to ruin their life financially.
You might wanna think again on that line of reasoning, because plenty of other countries have the same dynamics with respect to service employees, but they don't suffer the very US-only problem of ridiculous education and healthcare costs where calling an ambulance can ruin someones life.
My point is simply this: a person who helps individual customers in any industry isn't much more productive today than they were fifty years ago.
That may change, and it may benefit everyone except the people who get fulfillment in their life from the one-on-one human interaction they get from people who need services.
As a former first responder, I'm interested in hearing more about how AI-powered ambulance services would work. (related question: will the 911 dispatchers be AI?)
I don't think first responders are ever going to be at risk.
Administrators, on the other hand, are a massive part of the costs in the health sector (IIRC the Obama administration chickened out on truly reforming healthcare exactly because the number of administrators that would be made redundant would tank the economy). A significant amount of administrative work can be automated.
> It sucks for the employees, otoh it might be the only way we're going to beat Baumol's Cost Disease.
In the past few decades productivity has exploded, but service employees have largely failed to increase productivity in any way because it's harder to automate these tasks.
It's the reason the costs of things like education and healthcare are downright extortionate, the reason you're paying back your college well into your fifties, the reason you don't call an ambulance for someone in the US because you don't want to ruin their life financially.
Weird. I thought it was the fact that you have a cohort of people who are grossly overpaid to represent people who do none of the work yet expect an ever-increasing amount of value created by the work to be shifted to them every 90 days, no matter what, forever.
> We may have to trade the personal fulfillment in these jobs for the broader affordable access to these services.
Then you'll run into two problems:
1) no one will want to do necessary jobs without increased compensation, which is at the root of your analysis of "Baumol's Cost Disease"
2) at least in the US, you'll have a bunch of increasingly miserable people living in a society that gives them less and less to lose while increasing the availability of things that allow them to take out their frustrations upon themselves (substances) or others (weapons)
Yeah I definitely agree with this. AI has become very useful to me, but it has also definitely automated some of what used to be my favorite parts of my work.
I am having some success in working to acquire a taste for different parts of the work. But I suspect that this won't be an option for most people.
Most people don’t have jobs they enjoy, programmers have somehow escaped this; better pay and more “protected” time instead of doing what the rest of us were doing all along
They did till 2025. Now more than half don't enjoy it as the technical puzzles/plumbing is gone. The outcome oriented folks find it fun (and amplified up).
Where did the 20% number come from? I’d argue it’s way more than that (or variable, i.e. dependent on who’s using it/how it’s being used/what it’s being used on).
Having said that, the number, to me, doesn’t even matter. You could replace that with 200%, and it’d be just as true.
The vast majority of jobs are not full-filling or enjoyable. Because there were way more job seekers than jobs.
Programming was one of the ones which was, because there were fewer programmers than openings. Now that's flipping, thus naturally, the enjoyment is going to be sucked out of it.
anecdotal but a good number of people have clearly stopped working as much or quiet quit. some even left the company or the industry. CEOs are also telling everyone to 5x their output, AI will replace them.
Whatever productivity gains models are giving us is being eaten away by other factors.
For me it feels less like filling out reports, and more like mentoring an intern who can search for stuff really quickly but forgets everything at the end of the day due to anterograde amnesia.
Except the intern is trapped inside an iron lung and must communicate entirely by text. And also has zero real creativity or self-motivation.
> What's a 20% productivity gain if I constantly feel deflated by work that used to energize me? That's going to give back the productivity gain and more, while also decreasing my quality of life.
This has been the story of humanity since the industrial revolution.
I mean that sucks, but I have many tales of people, who were passionate and outstanding in what they did, and were rewarded with a leadership position for their efforts.
Now they get to fill out excel sheets, babysit people and sit in planning meetings.
Many successful people in their 40s could quit work already if they only decreased their monthly costs.
I think it is more important than ever to manage your wealth in a way that sustains you from capital alone in a world where employment gets progressively more toxic.
The way to achieve it is buying maintenance efficient and cheap car, make renovations smart, make good choices all around to minimise expenses.
Operate your life like a corporation. First, cut the expenditures.
For me the ability to do whatever I am interested in at the moment, is worth almost any sacrifice.
Then I can seek one time contracts or short time jobs that fulfil my mental needs.
My monthly expenses are no more than 5000 dollars and mostly consistently less than that.
Which is okay because the money spending doesn’t bring me any joy nowadays. The money isn’t what gives me happiness. Only other people can provide that and activities that are dirt cheap usually, like reading or broadly understood hacking
Consumption is a short lived and deceptive joy that causes more guilt than whatever dopamine it is worth really. Governments hate people like me which means I must be doing something right.
I think that's a very simplistic take, it doesn't account for people who need to support children's education and/or healthcare for parents.
Both are areas where costs are going up at a rate which is much higher than official inflation rates, and comp increases often don't even match inflation!
So the need for working till you're forced to give up still remains.
Also disregard this because I don't live in USA with its ridiculous healthcare costs even if my conversion of numbers accounted for purchasing power parity, it did not calculate monthly healthcare expenses that in my locale are only 1/7 of median national wage per month flat which I assume is a much better ratio than in USA and very predictable.
As per children, contraception exists and you should use it as kid is a dubious and too risky investment. You could for instance raise a little sociopath or two.
Healthcare for parents is for me at least letting them into a boat in the ocean when they get first alzheimer's symptoms like in Hemingway Old man and the sea. Native Americans used to do it. They just released their old guys to the forest or steppes. Which reinforces my point about the kids being rather bad investment. Do not count on that.
It's objectively more wise to save 500k-1mil from raising a kid to invest that money and spend it when old, hiring a professional caretaker. Have friends, spouse, not kids.
In any case it probably only matters that USA healthcare costs are so exorbitant. This is the crux to why it could never work for americans unless you are at much higher financial level than any sort of frugal upper middle class specialist in 40s.
It's the only western country where retiring earlier is so expensive. I thought maybe it would make up in the extra money you can make and save in taxes without VAT but no, after research I think that healthcare is so much pricier that it is incomparable.
So it is not simplistic but it's different environment altogether, one in which it actually works and I am really grateful for that.
Most people don't have the luxury of finding joy and meaning in their work. You aren't hired to have fun, you're hired to create value and wealth for your employer. Just do what literally everyone else does and grind through it until you get a pension and hope it's enough to let you die with a bit of dignity.
Here in stupid town we stab ourselves with knives. It hurts and we get infections but everybody does it. We don't even remember why we do it any more but we do so you have to as well.
The real problem is the amount of value that gets left on the table.
Also, mental health is just as much a part of health and well-being as the physical.
We know why we do it. We do it because we live under a system that requires most of us to bleed for our supper. And now tech is being normalized to work like every other job, and techies get to feel the knife. A lot of people are going to be shocked to realize what their actual relationship to capital has always been.
If we know, why would a lot of us be shocked? Seems contradictory.
That's probably too much a troll so while I get that you are grinding your anti-capitalism axe, you're also seeding hopelessness and trapped falsely fatalistic thinking. It has gone this way but it doesn't have to and the whole arrangement is a local optimum with much higher global optima available.
In fact there are many of us who have created a parallel life and extend ladders all the time (within our capacity) for others to join us.
> That's going to give back the productivity gain and more, while also decreasing my quality of life.
Unfortunately though, what does that matter? Your employer does not care how do you feel. You are paid to bring them benefits, they aren't running a charity. If you do feel down, that is your issue and you shouldn't let that influence how you work.
Just to be clear, I don't like that either. But it is what it is.
Mental health is health. Negative mental health impacts are particularly significant when your job is to perform mental labor.
It's not defiance, it's reality: if you drag down my motivation, that's going to drag down my productivity. I don't really have much active control over that.
Driving everyone's QoL to be as bad as possible will lead to increasing enshittiffication in the entire market.
Consumers will be spoiled for choice between deeply mediocre options.
Besides, what's the point of adopting new technologies if it's not to increase the quality of life? If everyone just exists in service of the product development lifecycle, who and what are the products actually for?
In my experience, for every customer support agent that really wants to help people and cares about their problems, there are at least ten who don't even read what you wrote and answer with prefabricated blocks of text that have little to do with what you asked. If AI customer support actually tries to understand what I ask of it and help me, and there are still (motivated) humans available for the more tricky cases that AI can't handle, that might be a win...
6 hours a week is low, unless its the average spread across industries. I think I spend more time in Claude Code via the CLI versus any other app I have on my laptop.
Like others said, the frustration is when it gets something so wrong you just think "wow, how'd you mess that up?" but when it gets it right its kind of nice. I also dont like that I basically tell Claude what to do, and then either go to busy work or waste time on the internet.
I kind of enjoy exploring black boxes, trying how different inputs are mapping to differences in outputs. It's kind of like hacking. The problem is, they keep altering the box.
It feels like Greek mythology should have some metaphor for "apparently simple structure that is so complex it leads anybody that studies it into madness". But I can't think of any name to put there.
The third sentence got to what my objection was going to be. It's fun trying to make the thing do what you want it to do! That's why many of us like computers. It's the randomness that sucks and makes the process unsatisfying.
Working with AI is trying to reduce the probability it'll pick undesirable paths. It's an exercise in trying to avoid what you DON'T want.
I suppose it's the same as asking someone else to take care of a feature and hoping they understand what you have in mind. The difference is that there's a lot of context that's shared between you and a human developer that is simply absent with AI.
I wonder how many hours of old stye non-overhead that is. And by overhead I mean mostly meetings. Then again maybe generating paperwork is included so that eats bit into it.
I theorise that many social ills come from workers having less pride in their skills and achievements, and a greater sense of social alienation, due to automation.
If you spend countless hours at work, and you partially define yourself by your work, and you realise you are easily replaceable then I cannot imagine this comes without mass social malaise that manifests itself elsewhere.
When you know you're essentially babysitting the workhorse to ensure it doesn't go off the rails, I can't see job satisfaction, and the social consequences of such, increasing.
> When you know you're essentially babysitting the workhorse to ensure it doesn't go off the rails
I don't have data to support this (other than, I guess, my LinkedIn feed), but my impression is that the management class is pushing AI _way_ harder than the worker / craftsperson class.
And if that's true, I think it's perhaps because it's something they understand: you tell AI to do something, and (with varying degrees of success and less complaining)... it does that thing.
To the extent that I've seen craftspeople adopt AI, it's been because they recognize its usefulness as a tool to further their craft. I don't meet many craftspeople that enjoy watching any[one|thing] do their work for them.
Does that mean the Human element can be condensed to a single Manager with the right skills? A Director above them? Is the VP above them directing the agents?
Is this another variation of Conway's law, where orgs design systems that copy their own communication structure? Seems like that is how my Manager/Director approach it. Then again, they are making slide shows and obviously AI assisted reports, not something that needs to be stable and responsive for the entire product lifetime.
But to your point, the manager sees it as a structure to manage to increase productivity. The craftsman sees it as a tool to further their craft. Each is driven by a different methodology and use case. Can that mesh unless specifically directed to throw AI at no joy work?
> I theorise that many social ills come from workers having less pride in their skills and achievements, and a greater sense of social alienation, due to automation.
I agree entirely. Even in an idealistic fully egalitarian post-scarcity society, to truly be happy I think most people would need to do work that they can feel a sense of accomplishment about. The problem is that work at most jobs is increasingly just toil. Any possibility to scrape some tiny flakes of satisfaction out of the toil gets removed, often for no good reason.
There's been a fair amount of research done on this, and while Marx's alienation theory isn't totally out there, what makes the most difference for workers is autonomy and purpose. We need to feel like we have some control over what we do, and like what we do is meaningful.
That's easier with things like farming, but it's totally possible with highly-automated jobs; you just can't have managers treating us like we're machines.
I see some similarity to how I felt when library management/wrangling became a huge part of software development.
In the last century I enjoyed crafting my own 'libraries' of functions that I could then use on the projects I worked on. As time went on, there was less and less of a point doing that as the odds rose near to 100% that there was 'a library for that' thing I was working on, so I was encouraged/forced to download it and use it.
It solved problems and was quicker than writing bespoke code (and libraries were hardly a new idea), so the logic was hard to deny, but I enjoyed my job less over time. Now I've risen up the ranks and now code mostly for fun (yes, I use AI to write functions for me) I look at what it must be like to enter the industry and think it all looks very different to how it did when I started.
You could argue that AI has done this much faster than it did in my early career, so people have less of the 'boiling frog' experience I had, and more of a 'sudden shock' to the system.
It's sad, but I've been doing this to other industries all my career, so I can hardly complain.
You pay per token, even on subscription models the limit is tokens.
If I was valued at 1 trillion dollars, and I was in the hole enough to sink a couple small countries' GDP, maybe I would slowly start to optimize to maximize token usage.
I want to sell tokens, how do I sell more tokens? Not by doing the same work in less tokens, that's for sure.
This is like if you pay me by the hour and then excitedly tell me that you keep paying 10k a month and it's great. I will most certainly not work faster, in this hypothetical, if you tell me you love spending money because it gives you a dopamine rush. I would probably spend a couple more hours REALLY thinking about the task, maybe writing some docs nobody will read, maybe considering multiple options, doing benchmarks, doing research, and then later maybe ill do the actual task as well.
Im not saying these AI companies are scamming us, but the incentives are there and extremely clear. The only thing currently holding it back is that there is some vague kind of competition.
Anthropic just said they're going to limit the ability of the new model to assist with building foundational models.
The ability to assess the user's goal and tune the inference accordingly is there. So, what methods are there to tell if a provider has their thumb on the scale?
Plodding along a longer path to the same goal would burn more tokens without necessarily a decrease in solution quality. Maybe a few more docs.
It'd probably still be worth it to make tokens cheaper. Jevons paradox [1] seems to apply here, where making compute cheaper unlocks use cases that were previously priced out, increasing overall demand.
I spend at least 6 hours a week arguing with bots owned by other teams, as I’m unable to reach a human before I bypass their bot. 10k person company, clients are paying for my time.
They are, eng makes 600k/yr and I make 2, so their times is worth triple mine. It’s cost effective to make me toil and drown in frustration if it reduces ticket volume.
Actually having to work and take responsibility is extremely exhausting compared to corp life, I experienced both, I currently have no corp job, am responsible for a lot of apps and hate every part of it.
Boreout is so much better than burnout and I never worked a job that allowed me to have none of them.
I've found that setting good guardrails, and running in a sandbox so that the agent doesn't keep asking tedious permission questions, makes things go a LOT smoother.
Generally, I spend anywhere between 15 mins and an hour setting things up (depending on how well the project is set up for AI work), and then set the agent going, coming back in a half-hour to an hour to check its progress. Generally, the tooling keeps it honest (for golang, forbidigo is AWESOME). 80% of the questions the agent asks me require a lot of thought. 20% of what it does needs correction.
The other thing to remember with LLMs is that they are NOT human, and won't react in a human way. So you'll see strikes of "brilliance" followed by the absolutely bizarre. But good guardrails keep that to a minimum.
> sandbox so that the agent doesn't keep asking tedious permission questions
> 80% of the questions the agent asks me require a lot of thought. 20% of what it does needs correction.
I've found even the permissions questions give me veto power over fruitless lines of exploration, especially in planning mode. For instance, it wants to use tools I don't have installed to access information that I have made available elsewhere? I get a chance to override this decision by declining the permissions check and redirecting it. Feels tedious, but helps me understand what information sources are influencing it. I head off a lot of bugs this way.
I never let it go into planning mode, other than to output a plan file that I can audit before giving it the go-ahead to implement. After that I don't want to be bothered, so --dangerously-skip-permissions keeps all but real questions out of the loop, and I can do something else while it works rather than babysit.
This is my impression too. Whenever it needs permissions outside of a small set of defaults I've allowed, it's often because it's trying to do something ridiculous that it doesn't need to do.
I think the yoloist counter-argument is "So what? Let it. It'll take longer that way and consume more tokens, but you can work on something else in parallel instead of being hooked in to this one session".
I do it with every project I go into. First step is setting up the documentation so that the agent can navigate it quickly, knows the idioms, knows the test gating procedure, design principles, coding standards, testing policies, etc.
Once that's set up, I spend time laying out the planning for whatever feature or fix is being worked on. For fixes the agent is pretty quick and usually needs little guidance. For new features it's best to have more of a hand on the tiller.
AI should be assisting us, instead it's doing the job and it's us being an assistant to it. This is a monumental shift that people seem to be missing in how knowledge working is changing and it's going beyond mere coding.
Guardrails, prompts, whatever, it's us helping it doing the job, not the other way around.
Opus 4.6 was the last genuinely good assistant LLM, but since then it's quite clear that the training/reinforcement is focused "given prompt -> do task" so it's behavior is more and more about doing it itself, not helping you. If you try to use it as an assistant it just sucks and is perma wired into finding the solution. Many times I want it to help me investigate, and his answer will still be focused on the fix, not answering my questions.
4.7 first, 4.8 later and fable are absolute disasters as assistants.
Fable in particular is so "intelligent" that it will push with very strong and intelligent takes even if it is completely wrong.
Wow... Our experiences have been very different, then. I've found each upgrade of Opus to be a noticeable improvement in its complex reasoning and delegation capabilities over its predecessor.
To me, this feels in many ways like a technical manager or team lead's job, where I guide the process along using my knowledge and experience, and then let the agent fill in the rest (to the best of its ability).
The agent can't really learn from its mistakes (at least, not without consuming precious context), so I apply a blameless postmortem process, updating the guardrails whenever it goes astray in the same way more than once.
And really, I'd rather be contemplating the more difficult and interesting questions of architecture, environment, ergonomics and market fit, so it suits me fine.
> Wow... Our experiences have been very different, then. I've found each upgrade of Opus to be a noticeable improvement in its complex reasoning and delegation capabilities over its predecessor.
I haven't stated that it's not more capable nor more "intelligent", it's the opposite.
I will try to expand on what I mean.
LLMs "character/persona/tendencies" are increasingly less about acting as an assistant and more about finding the solution itself.
I use AI in a specific way: he assists, investigates and answers my question. I do the coding. It is increasingly difficult to use it as such, because it quickly jumps into giving me solutions instead of answering my specific questions.
I'll give you few examples.
I asked it to investigate DNS handling details in phoenix emailer module work, he did very little investigation and jumped into why I should've used magic links instead. Instead of assisting me in my research, it was hard wired to solve the problem (the wrong one, with a very wrong solution).
Today at work, I had a problem with batching, I wanted to understand if batching was even needed at all for our use case, and he kept circling around how to fix the batching bug instead. That's not what I asked it to do, yet, it jumped to the "solution".
I am increasingly frustrated by these models "personality" and tendencies that are unhelpful to assist me doing the task at hand and more on it doing it and me merely assisting/supervising.
Sure, very detailed prompting on how he has to act helps, but wait few turns and he drifts again to his default solution vomiting state.
Which makes me think that these models are hard wired on this mode of operation by consistent training and reinforcement of jumping from prompt to code solution.
Ah yes, the agents by default are very "implementation" oriented, which is why I instruct mine to never implement something without formulating a plan first for me to approve.
Another thing they tend to do is rely on their own context -> memories -> training data. And if that's wrong then they'll continue with it until you instruct them to research, after which they usually get the right answer.
I've noticed that the newer models keep track of what you type so as to anticipate what you're likely to say. For example, today Opus 4.8 said "You usually don't want me to commit until you've checked, so the change remains uncommitted."
I think this is just a misunderstanding of how most technology has always worked?
Consider what is happening in most construction sites. The heavy work is absolutely from the technology on site. But without people there to oversee it and keep it working, it would fail.
And that is almost certainly true at any industrial site. Indeed, look up videos of high tech looms. A large portion of the technology added to them are so that the operators can locate the fault and fix it.
The problem (okay, one of the problems) with renting other people's models is, as you mentioned, that they can and will change out the model without notifying you ahead of time, and you don't always get to control which model you use. (They might decide to retire it, and you won't be able to get it back if they do).
Which is why (well, part of why) I think the long-term trend will be towards self-hosting models. Right now the frontier models are far enough ahead of the self-hosted ones that there are lots of people willing to pay by the token to rent someone else's model, because they get more value for money from that than from self-hosting models.
But the frontier companies won't be able to keep up their current levels of expenditure forever. At some point the investors are going to say "Hey, so, um, when am I going to see some return on my investment?" and then the current subsidized subscriptions (including the one my employer uses) are going to go away, much like what happened with Copilot this month.
And then the locally-hosted models are going to suddenly look like a more attractive picture. Because where you might have been willing to spend $100/month/employee to rent time on models in someone else's data center, you might suddenly balk at spending $500/month/employee. You might say "Hey, you know what? A $50,000 up-front capital investment is only, what, one month's worth of subscriptions for our 100 employees? Yeah, okay, I'll approve the hardware purchase. Get that self-hosted model set up and then we'll cancel the subscription and switch over."
Not everyone is going to do that. But once the locally-hosted models are good enough, the first few people who do so and report success are going to start a snowball effect. And it will likely be driven by money first, but it will also have the effect, that people will slowly discover, of meaning that you can better predict the model you're using. It will continue to work the same way next year that it is working this year; or if it doesn't, it's because you chose to install the new version.
And when that happens (I'm saying "when", not "if" because although it might take some time, I think it's inevitable in the long run), the frontier-model rental companies are going to struggle to stay afloat. Except for the ones who saw this coming and transitioned to a non-subscription income source somehow (maybe by selling licenses to self-host their frontier models for $$BIGNUM), or who have some other revenue stream besides renting out models.
AI should be assisting us, instead it's doing the job and it's us being an assistant to it.
If you're a manager and you ask a report to do something and they come back with a question, does that mean you're now their assistant?
I give agents the tasks, I answer their questions, I make choices about the tradeoffs in their plan, I supervise their implementation, I review their output, I have them walk me through things. In what way is this not delegating to them and managing their work, just like a more junior employee?
Well... as a human software engineer, I've been the one with very strong, intelligent, completely wrong takes. The question is, are the LLMs improving faster than you can improve a junior dev? And is their ceiling as high?
Your experience pretty much mirrors my own. I hate to be the 'they're holding it wrong' guy but there's certainly a lot of people out there that have no real idea how to effectively leverage AI.
That’s a problem with the tool not the people. AI is marketed literally as writing one sentence and having some app perfectly output. Just check any of the landing pages for Claude code or codex or GitHub copilot…
No, it literally isn't. I just looked at the landing pages for Claude Code, Codex, Cursor, and Copilot, and literally none of them have anything about "writing one sentence and having some app perfectly output", or anything remotely like that. In fact, just the opposite: they all make clear that they're built for ongoing collaboration with AI, and have detailed descriptions of what that looks like. No one advertising the idea that you can one-shot perfect apps with these tools.
i've seen a number of articles claiming things like "devs self report they'er +x% more productive with AI, but actually they're -y% LESS efficient!". and i think that this is explanation for why.
as a boss (or researcher) i'm going to measure productivity based on amount of output per hour that i'm paying you; as a workers, i'm going to measure productivity based on amount of output relative to the amount of effort i'm putting in.
so what may be happening is that bosses see that output is at 80% (productivity down!) but workers see that they can give that 80% output with 40% effort (productivity up!).
Not sure among devs, but I do know that in other positions in typical corporate bureaucracy, people have a propensity to not report their own automations or productivity gains upward, because the reward structure isn't there.
Early on in my days as a sysadmin, I automated a ton of my role when the rest of the team was still doing ClickOps. The reward for doing so was more work and expectations without the additional pay increase to justify my new found productivity. That happens all over the workforce, and so people will just keep it to themselves. I learned my lesson at that first job real fast that if I'm able to have the same, or greater output, for half the time, I keep that to myself so I can use the automation to free up my own time instead of have it filled by the company.
I wonder how much of that is happening now with AI in non-technical roles.
> so what may be happening is that bosses see that output is at 80% (productivity down!)
If an initiative produces only 80% of the previous results and you’re paying large token bills on top of the same wages, the AI is going to get cut off.
> i've seen a number of articles claiming things like "devs self report they'er +x% more productive with AI, but actually they're -y% LESS efficient!".
Are you thinking of the old METR evals? Their more recent evals showed an actual performance improvement.
The old report is still circulated as bait for AI skeptics.
I think the old report you're referencing is this [1] from July 2025, but I can't find a new report. This [2] links to a new dataset at the bottom (that maybe shows improvements?) but it seems like they chose not to write it up because of perceived flaws in their study. Is there a more relevant report I'm missing?
> so what may be happening is that bosses see that output is at 80% (productivity down!) but workers see that they can give that 80% output with 40% effort (productivity up!).
So why is it that the bosses are the ones that are so enthusiastic about adoption?
Well the obvious is that, monkey-see-monkey-do, and they don't want to miss out. But the insidious externality is that they are the most likely ones (and higher ups) to be invested in nvidia and others, so when they push you to use, it creates a viscous cycle.
It's horrifying how many people I see here being threatened into using AI. I guess I'm fortunate in that I have a choice.
I've used it plenty, more recently than ever before, but I'm coming more and more to the conclusion I don't want anything to do with it.
Botsitting is low-skilled. Low-mental reward work will become low-financial reward work before long.
I was initially worried that I had to keep up, learn the new tools or get left behind. But I'm beginning to see it as an entirely different domain of work, and one I'm fine not doing.
I think I'm best off preserving my skills and autonomy, rather than fighting with an agent and fretting about tokens for the foreseeable. I've no desire to be a botsitter.
Worst case scenario, this really is the future, and I become unemployable in tech: I've done other sorts of work in the past, things I'm happy returning to if this is the future.
I've a hunch that's still going to be a while off. I hope so anyway.
i was a senior designer that basically had to reclass to a manager/senior manager role just to keep a job that didn't touch AI as much.
the wild thing is, i liked my job 3 years ago. i liked the satisfaction of creating for sales. they could be annoying, but it was minor normal (human) corporate squabbles. now, when they're explicitly bragging about getting rid of designers and treating us as sub-human, it's creating resentment and misery. personally job satisfaction sharply decreased in the last 6 months let alone the last 3 years - i'd say from 80% happy to 20%.
personally, i think suicides are probably going to be on the rise in a big way specifically due to this; and that's just the people who can hang onto a job right now; cannot tell you how many designers and animators are out of jobs and desperate.
My challenge has been trying to manage my higher-level context. I've gotten a pretty good setup where I have project-level orchestrator agents that can spin up workers to implement tasks with minimal oversight, and the resulting work is usually quite good (especially after I give it the mandatory "make the comments less verbose" refining, etc.). But that means I'm doing even more context-switching. I've gotten to the point where I have a half-dozen draft PRs that just need my review before I tag my colleagues, and trying to dig up the context from all of those tasks can be paralyzing.
This is something I am very curious to learn from others since I am, relatively speaking, quite new to HN.
HN posts seem to go from very supportive to very critical of AI with very few articles striking a balance.
Was there a time or technology phase in the past (I'm thinking perhaps when everyone was going over to GraphQL just because of the hype) that you recall being similar to this? If so, how did you maintain curiosity without feeling yourself go through the extremes of "This is the worst / best thing to ever happen"?
HN has a huge user base. If the article is pro-AI, you'll see top comments praising AI. If the article is anti-AI, you'll see top comments deriding AI. The article and headline is just an attractor to those whose opinion is aligned. It's not unique to AI, basically every thread has your typical nerd-fights, but AI has been particularly polarizing for many reasons.
True but it also depends on the initial audience that sets the tone, and the time zone. In my empirical experience, pro-AI posts fare much better when West Coast is awake, and response is more muted when Europe is awake.
I just started using Claude Code for my work as a sysadmin. For my work, it's great. I don't need to wrestle with MySQL joins, claude gets even the most complex ones right WAY faster than I would. Same with new Terraform stuff. Things that would have taken me a day are cut to less than an hour.
So for my work, it's made me much better at my job. Much faster and more accurate.
I can write a simple query before Claude finishes reading, querying the semantic layer, checking my files, then writes a query that I have to approve, reads the results, hides them (ctrl+o usually works), and gives me a summary.
We’ve reached this inflection point where it’s faster for me to do most tasks again.
I’m sure fast mode costing more money plays a role.
Personally I find LLMs absolutely terrible at writing Terraform and full of hallucinations. But also Terraform is my bread and butter and our use/workflow is fairly advanced. And we're multi-cloud + baremetal. That wasn't an area where I was going to get a ton of value out of LLMs anyway.
I don't see a lot of talk about how AI development breaks the old feedback loop of write code, watch it run, change it, repeat. I really hate sitting around waiting for the agent to get done planning, reading the plan, then waiting for the agent to get done coding. It's those 5-10 minute windows when its working that really sap my patience and suck all the fun out of our jobs. Writing code by hand is just more fun.
This is something that I don't see discussed a lot in these conversations, but its true for a ton of folks.
I didn't end up with a career in tech because I wanted to tell a bot to do the fun part of my job for me, leaving me only with the boring tedious parts. I didn't sign up to be a full time code reviewer, and I certainly never wanted to be a manager, yet alone a manager of bots.
It also can't help but spark feelings of "Why am I getting paid 6 figures for this??" and that makes me nervous for the future.
I imagine the engineers and assemblers in factories pre-assembly line felt the same when things started getting automated there. There's an element of craftsmanship that gets taken away as the product moves from being artisanal, hand crafted to mass produced.
I wonder if its too late for me to pivot to hardware
Yeah its hard to deny just the raw throughput from the AI. Like it really is doing work in hours that would take me days.
But those times when I had to drop down into a repl and play around with the output of a method. Or try different ways of doing what anyone else would think is boring, like array manipulation - that's a lot of what I actually LIKE to do.
A big part of me just hopes I can hang in there for another... decade, or two. Then I can retire! Maybe.
I think a lot of developers who enjoy writing code are in an internal battle knowing they could write it themselves or AI could do it and probably do at same speed or many times faster. It's a contradiction as a developer because we always want to do whatever it is we are doing at the optimal speeds but we also like writing code. So, you're being forced to choose where you stand at a fundamental level.
Absolutely. And I'm saying that the fix is to decide to write code. The AI isn't faster at producing good quality, well designed code that you'd write (if you're quite competent, which anyone can become with practice).
The AI code generation speed is a lie, it doesn't get the same work done in less time, most of the time it gives you a different version. Thinking through the problem yourself can yield better results for the overall product, in my experience.
I am actively developing on 4 projects fulltime. Kind of like double or triple dipping but just 1 employer. With a Claude Code license I'm running 2-4 CLI instances with AI. I'm botsitting 6-8 hours a day and getting more done then every. I love how much I can build using AI
This kind of reminds me of an article that I saw on HN ages back, there's like a subset of office workers who automated their Excel jobs, and just show up to work, read books, and do literally anything, while Excel does their work for them, and they collect their paycheck.
I don't know what they're complaining about. AI has freed us from the drudgery of craftsmanship, letting us focus on the important stuff—managerial and administrative work!
(There's a reason why I call it the MBA's stone. It transmutes all knowledge work into a problem of management.)
For me, AI can sometimes create a false sense of productivity. It's similar to how in the past, people would spend time creating the perfect setup with notion templates, pomodoro timers and productivity tools, or tweaking their environment for maximum productivity, instead of actually doing productive work.
But now it's happening at the company level: "We're going to add a chatbot to increase productivity! Now MCP tools! Then agentic workflows! We’ll add skills, and now productivity will go up! Maybe loops will do it?"
Understanding what is going on with AI productivity is … frustrating to say the least.
The best I can say is that genAI is a self reported a 20% efficiency boost, and for a very (very) small group of people, it’s maybe a 2-3x boost. (And if you are at a frontier lab, you go fly into the big bucket of exceptions)
At this point, for most use cases, AI productivity is either the equivalent of giving people 3D printers, and seeing little benefit, or signing up for an outsourcing service, just without the development of human capital anywhere.
>Understanding what is going on with AI productivity is … frustrating to say the least.
Agreed. I think one of the hardest things about it is that productivity != value. You can push all the code you want, but if it's not driving revenue up or cost down, it doesn't matter economically.
Here is the best data I've been able to find. An observational study of 4000 teams over 2 years across many different organizations. Data gathered from their task management, version control, and CI/CD tooling. Critically - this is not survey data. It's much more direct measurement.
I think it depends on how you measure the boost. If you are talking about generating a first draft then yes, the boost is there. If you’re talking about completing the project in all well tested and architected aspects, then overall there really isn’t a boost.
6 hours of debugging and docs reading is not equal to 6 hours of prompt fiddling. The return of value beyond the few fixes applied will be almost nil from the fiddling.
Yeah, Amazon warehouses are just the same. Humans are only used for tasks beyond the comprehension or physical ability of a machine at that point in time.
The problem is, we haven't had the debate on a societal level if we want to go the star trek route (aka, we give our darn best to automate everything so that humans have the time to do whatever they want) or the realcommunism route (we ward off automation so that we have jobs for people).
The result of that debate not having been made is the third possible outcome - rabid capitalism automates everything as soon as it is profitable and lays off the humans, focusing on getting higher margins out of less people if need be; the best example for that IMHO is Disneyland or Vegas going on ridiculous nickel-and-diming tours. In the end however, there will be no one left any more who has employment and we'll be in for quite the riots.
I could care less about bot sitting (haven’t we always written our own automation?), but it’s botsitting the unverified slop that people send you that fuels frustration. I thought I worked with competent people who respected me
Our product lead/manager recently sent me an AI generated PRD (complete with a Claude Code spec!) to build a core feature which we have had for over 2 years (and is a highly used core feature by our customers).
I just can't imagine tanking my trust with my coworkers by doing something like that.
So we're now in this world where everyone is instantly 10x more productive at turning their thoughts into code. Now, think about the coworkers you've had that are middling to mediocre. Do you want them to have a tool that makes them 10x more productive?
That's what I wonder about, what happens to all those folks.
Your coworkers haven't changed. What changed is that people can hand off work they never had to think through themselves. So you don't know what they checked and you don't know what you need to. You just have to read the whole thing.
Guys you are privileged to do a job that you love. The privilege is even higher when the things you love aren’t even needed anymore. If this is the case you are on the verge of losing your job. Key word is verge. Not yet, but the trendlines point to it as a likely possibility.
I have barely technical people I knew from college pushing the latest AI certification (I can't imagine this staying relevant for longer than 6 months) and one being hired as a frontier AI engineer.
It reminds me of 2000 all over again where knowing html got you a six-figure job.
This really hit home for me:
In some cases, workers are also being asked to automate the parts of their jobs they enjoy most, Hinds said on the podcast, pointing to customer-service employees who enjoy building relationships but are increasingly expected to supervise AI agents instead.
"That's what gives you joy and meaning at work," she said. "That is very dangerous."
What's a 20% productivity gain if I constantly feel deflated by work that used to energize me? That's going to give back the productivity gain and more, while also decreasing my quality of life.
This is an important point. My light-bulb moment was when I talked to a product owner in a previous job, and I expressed surprise around an expensive planned change, because it didn't seem that valuable to our customers.
He said, "Almost half of what we do is not that valuable to our customers, but it's valuable to him, and her, and him", pointing through the conference-room window at my fellow programmers, "and that's why we do it. If we only did things that were very valuable to our customers, we wouldn't have nearly as many good engineers on the team as we do."
Do the engineers not derive enjoyment in their jobs from making the customer experience better? This does track with what I've experience in my career, where we've gone from everything being to better the user experience to tech companies sort of trying to out-do each other in their technical solutions while the software continuously gets worse and more antagonistic.
Plenty of value comes from things the customer doesn't care about.
Customers want features as fast and as cheap as possible. I derive joy from solid test suites that avoid me getting paged while on call and team processes that don't allow config changes on Friday so pages don't happen on the weekend.
Very few craftspeople derive their joy from the customer experience. An electrician isn't happy because their work allows me to watch TV. A carpenter isn't happy because a new set of stairs lets me get to the basement faster. They're happy because of their perception of the quality of their work. This goes away when the visible or fun parts are no longer "their work"
I wouldn't agree with that. The issue with software is that the people you make things for are usually anonymous and you'll never meet them, but if you've ever built software that helped someone and you witnessed it, it feels really good.
Meeting your customers feels good. It's still not why I do it. I've written components that were used by billions of people and generally loved (pieces of the Windows 7 taskbar) and talked to people about that work. I've worked in the education space and talked directly to the staff at schools who use my product to hear their use cases, thoughts, and feedback. It's fun, humbling, and rewarding. It's not what motivates me day to day.
how are you navigating the AI wave? Those parts of the job are fully out, and the customer engagement/value are fully in.
I don't agree with your statement. Individual productivity is very different from team productivity. AI is much like the traditional cowboy coder - prioritizing speed above all else leads to disaster.
As to customers, customers value stability. They don't want the UI changing every week, they don't want random changes of the data model that corrupt their content, and so the traditions of quality, strong testing and integration pipelines, and so on are just as important in established products - in some ways more important than ever, as there are plenty of opportunities for your competitors to burn user trust and drive them away.
I have yet to encounter an established product demonstrating a massive improvement in velocity due to AI. Being able to spit out prototypes doesn't matter all that much.
I'm handling the AI wave by holding on to my sanity and skills and working to establish a reputation in preparation for big paychecks fixing the messes everyone else is creating with their tokens.
I don't see any indication that's the world that's coming, and lot of opposing evidence. But perhaps it's a bet you're choosing :thumbs up: .
This is the danger of isolating engineering from customers, or even internal customer-interfacing employees.
If all they see is code, they will get satisfaction from tidy code, not user happiness. One good thing about AI is it elevates product engineers because they more directly bridge the customer-product-code divide.
That's my take as well... the dev effectively takes on QA/QC and PM roles as a team working with AI for the baseline of development work. Of course, this is also a slightly different skill set and cognitive load. It also needs to completely upend how project planning happens when you are using Dev+AI in coordination.
Not just isolating them from customers but also from engineering-adjacent work that isn't code.
I've been at a place that is basically microservices slop (several dozen services per engineer). They're all poorly maintained and at least a solid 40% of all this code that they've written could have been just a traefik or nginx configuration/container.
When you have a lot of inexperienced (relative to industry) and overworked software engineers, the solution to every problem becomes to write code and writing new code should be a last resort.
Worse still, there's just a poor general understanding of the internet protocols they're working with and of how to do distributed systems right. Unfortunately with LLMs I've been seeing this get worse, not better.
They use the LLMs for code generation but not architecture review. Bad ideas are getting fully-baked quickly before anyone with good sense can intervene.
There are plenty of projects that are green lit that have good intent but are bone headed when it comes to solution and implementation. Good engineers hate these types of projects. Good PMs try to avoid these at all costs but sometimes your hand gets forced because some VIP, either internal or external volun-tells you to do it.
> Do the engineers not derive enjoyment in their jobs from making the customer experience better?
Quite a few don't, no.
Different people derive enjoyment in different things and some of the best engineers do not find satisfaction in "delivering better customer experience" but in working with, and improving, cool technology. Its up to management to find areas of the business where they can deploy these people in a way that dove-tails with business success.
Its also the case that only working on projects that "deliver customer value", and having to justify every single endeavor through that lense, is how you end up in a local maxima in your tech stack, get mired in technical debt, and then get lapped by your competitors who have the foresight to work on foundational technology that enables future velocity.
To be frank, its endlessly frustrating that your median Hacker News poster doesn't get this, and instead prefer to brow-beat people about how they're caring about the wrong things.
Platform and internal dev teams have customers as well. I'm not terribly frustrated that you don't get this. I've certainly worked with devs (and managers) who wanted to push new technology for the sake of using new technology, and they should have found a side project as an outlet for this.
The person you're responding to never said those teams didn't have customers.
It's not about new technology for the sake of new technology, it's about taking pride in one's work and what that person created.
Honestly, the American obsession with "everyone should think of what the customer wants" is exhausting verging on toxic. The people who talk about that point loudest are inevitably owners saying "you should all care more to make me richer". If you want your employees to care about the customer more than their own personal satisfaction, give them significant equity and significant autonomy such that they can see how their actions have direct impact. Saying they should think of the customer and then treating the employees as an interchangeable cost to be minimized is insulting and won't lead to anyone focusing on the customer.
> The person you're responding to never said those teams didn't have customers.
> It's not about new technology for the sake of new technology, it's about taking pride in one's work what what that person created.
Thank god, I really want to say that I appreciate that you got what I said. A simple upvote didn't feel like enough.
Employees are humans, not robots. Its inconvenient, but if you want a world-class team then you're going to have to deal with the fact that people derive satisfaction from different things, and you're not going to be able to motivate them by beating them into submission about what they "should care about." This may involve having to think creatively about how to manage your people instead of treating them as fungible work units.
I'm going to apologize in advance for being long-winded, but I feel there's a lot to unpack here.
> Platform and internal dev teams have customers as well. I'm not terribly frustrated that you don't get this.
Respectfully, this response here is a perfect example of you not getting it and assuming that I am the one that doesn't get it. You have not said anything that I have not already heard and understood before. The fact that people have different values does not mean they "don't get it." But saying something like "Do the engineers not derive enjoyment in their jobs from making the customer experience better?" does imply that you don't understand other peoples' values.
The fact that you posted "Platform and internal dev teams have customers as well." indicates to me that you missed the point. Whether they are on those teams and whether they consider other engineers their customer is besides the point; they may not derive satisfaction from "delivering value" to those people regardless. That doesn't mean they don't care about their customers, which is the take away the median HN poster takes, but rather that they are not energized and motivated at the end of the day by delivering value to them.
> I've certainly worked with devs (and managers) who wanted to push new technology for the sake of using new technology,
Sure, everyone has. But the flip-side of this is a class of people who assume any tech improvement that doesn't directly move a metric is just an effort at resume-building. Just as often I've seen efforts and building a more robust system as unneeded resume building despite clear need (usually because the need is very hard to measure).
> and they should have found a side project as an outlet for this.
I mean, this is incredibly dismissive and exactly the attitude I was talking about. No-one is saying that engineers should be allowed to just do whatever to have fun. Work is work. But ideally you find ways to organize your team so that everyone is motivated and energized by their work, and doing so requires that you understand that not everyone is motivated by the same thing. But in these discussions, the attitude comes across as "everyone should be motivated by delivering good customer experience and if they aren't we shouldn't care."
If there's no opportunity to give these sorts of people fulfilling work, then fair enough. It *is* work. But the attitude displayed here is that we shouldn't even try and understand their values and think about ways to productively deploy that.
As an aside about customers, internal and external customers are, in my experience, treated vastly differently. We care about experience for external customers, but internal customers is usually all about velocity and trade offs. The bar is substantially lower, and rough edges are almost always ignored. So I am skeptical at the idea that we can just frame internal users as customers and all the discrepancies go away.
It also misses the fact that other people on my team are also my customers, because they have to maintain the system! And I am also my own customer, because I also have to maintain it!
Force me to click hundreds of buttons per release and I'm going to be disinclined to go through that. You wouldn't have a surgeon have to go hunting for the right tools.
I dunno about everyone, but I do. It drives me up the wall when software is slow, stutters, etc. There’s almost no good excuse for it. But it happens anyway because of some combination of, “nobody cares,” and, “money.”
I derive joy out of people using a thing i helped design and they never get frustrated with it. It works, it’s fast, it solves the problem and it’s rarely a pain to maintain or extend for me or my team. Thats good, solid stuff right there.
That's kinda software development ... generally and not really new as far as I'm concerned.
Even everyone earnestly trying to identify the feature that IS IMPORTANT to clients and will actually be used and appreciated frequently is something of a crap shoot.
A better way to put it is that these things do have value to the customer, the customer just doesn't have a way to understand how the work you're doing provides value because it's the part of the product you don't see. If you clean up technical debt, improve test coverage, improve your deployment systems, etc, it doesn't change the immediate customer experience in a meaningful way, but it does allow you to deliver the changes that customers do see faster and with fewer risks.
This quote makes it seem like the work is self-indulgent, and I have seen that happen sometimes, but it's not half of what we do.
By that logic, pair programming should have taken off
I always found pair programming a bit.. hellish. Chatting about things. Rubberducking. Playing code review ping pong. All great. But the feeling of being shoulder surfed killed my ability to play around with the freedom that I do when I'm alone -- and that playing around sometimes led to better/more interesting outcomes than I'd have gotten otherwise.
Yeah, I always bought the premise that it was good for the software, but it didn't work for me as a person, it drained my energy way too fast to spend hours of the day having to be "on" in conversation with another person.
In many ways, I think what's working for me with AI is that it is very similar to pair programming, but without the social-emotional investment required to interact with another person for long periods of time.
thank God, if anyone knew how much time i spend swearing at our codebase i'd have a meeting with HR scheduled within a week
Far too many introverts in software development for that.
You can be extroverted and not want to spend an entire work day interacting with people without a moment of privacy or introspection.
Certainly puts "good" into perspective
Perhaps those “good engineers” need a reminder of who enables their situation to exist at all.
How this goes is the engineers start raising arguments in meetings full of nerdy technical terms such as “refactor”, “technical debt” and “accidental complexity”. As time progresses, more of them say more of this. At some point management learns that when engineers say engineery words like this you sometimes gotta say yes and let them do it even if it means urgent features are delayed, or the engineers will walk.
Good engineers would communicate effectively of something is impacting the product. Good product understands how to judge that impact and prioritize it. Frankly, if the engineers are making enough tech debt that they don't want to play in their own sandbox as part of regular practice, maybe they should walk.
Let them walk. Good programmers can write code that doesn't degrade the customer experience.
In the last 15 years it's been all about the programmer's experience and user-facing software is objectively slower, buggier, and more resource intensive.
It's a two ways street though, isn't it? With lazy/unskilled/unmotivated developers you will have a shitty product, or maybe your SaaS would explode the day it's starting to get traction.
Are you all hiring?
I mean if half of what you do isn’t delivering customer value maybe you don’t need to have as many good engineers on the team as you do.
When I was given a semi-ultimatum "use AI or get fired" kind of thing for writing code I had a brief bout of depression/sadness. Whereas my friend doesn't care/says "I get paid to not work". I have gotten past it, now I'm just like, I'll do what I need to do to get paid since unfortunately I'm in a lot of debt so I need this job. I learned to code in 2013 so I like typing the code myself but now it seems like a waste of time. I still write my own code for myself/hardware hobby.
If you're anything like me, and you stay in a coding/dev/IC track for your career because you like the work, you will eventually hit a point where you start thinking it's all meaningless. This happened before AI for me, but AI certainly reinforces it.
You come to a point where you realize that you're not doing anything that creative, or nothing you haven't done hundreds of times before, maybe every few years you switch to whatever new tech stack has gotten popular, but it's fundamentally all the same. And you start to realize that everything you do has a lifespan of a few years, and then you (or probably someone else) will re-do it.
As retirement starts feeling like it is something that will happen sooner than later, you look back and see that almost nothing you've built is still in use, or will be for very long after you're gone.
I hope to retire in about two years. At that point, I plan to not be using any technology or computers in my life for a while, or as little as possible. Maybe at some point I'll rediscover some of the fun I used to have writing programs for myself, but I suspect I'll need a long break before that happens.
> maybe every few years you switch to whatever new tech stack has gotten popular, but it's fundamentally all the same
So true!
But it's interesting that, from the perspective of someone in the middle, neither near the beginning or end of my career, I am (now, after a period of sadness) experiencing AI as a reinvigoration of fun in the work. But it's a very different kind of fun. I had totally lost the fun of clean code and figuring out new technologies and approaches and abstractions, just like you describe.
But now I'm experiencing the joy of thinking about what I can build, now that it's so much faster and easier to try ideas. I think this is actually getting back to an earlier version of my joy with computers. I can (vaguely) remember in my early years being like "wow! cool! I can make stuff that shows up on a computer screen!". But then it turned out to be ... pretty damn hard to actually do that, which led me to more excitement about all the ideas and technologies and techniques for managing the complexity of software engineering. But then that started feeling more tedious and samey, but I still had to put lots of time into it, there wasn't any other option.
But now all that is so much easier, and I'm rediscovering the fun of "wow cool, I can make things!", but now also with the whole benefit of the time I have spent doing the work of software engineering.
> you will eventually hit a point where you start thinking it's all meaningless ... You come to a point where you realize that you're not doing anything that creative
I hit that point after about 5 years. At the time, software development was mostly desktop applications (which are sadly defunct), and UI frequently doesn't require much in the way of algorithms. Then I focused on the challenge of design a UI that matches the user's mental model, which lasted about five years. After another a stint in mobile games and apps, I started doing contracting. The key for me was the concept of serving someone else' vision (instead of my own vision of FIRE), and some HN conversation on some of Steve Job's videos on work. This all boiled down into an attitude of "I will solve your problem (and you will pay me)" (from Job's discussion with the guy who designed the NeXT logo).
As a contractor, sometimes I solve your problem with a cool algorithm, sometimes by simply implementing your pixel-perfect design, and sometimes with a true-Agile back-and-forth of vague design into concrete software. Since I switching to the "I solve your problem" perspective, I haven't had any problems with feeling frustrated at just being a code monkey implementing stuff I don't need to think about. (However, I have discovered that I get really bored doing server stuff, and it doesn't save me from that. I just like something concrete, and not having to debug by putting in print statements and wading through thousands of lines of output, with only the reward that the variable contents is now correct. Turns out I want to see something. I get unreasonably excited about going from a blank screen and slowly populating it with widgets.)
>you look back and see that almost nothing you've built is still in use, or will be for very long after you're gone.
Software development has more in common with agriculture than architecture. The code always needs maintenance.
I've definitely found what you're describing at bigger companies, but I also previously had experience writing software at smaller non-technology companies.
Legal marketing specifically. Weirdly, my work had more impact, respect and longevity there than the place where I'm a much more senior engineer supposedly directing the work of a whole organization of engineers. I had it better where I was a 1 of 2 than a leader among hundreds.
Small successful companies are great, but the hardest thing for me psychologically has been when I'm at a small company that is struggling to convince anyone to use its product. Being a small cog in a giant machine serving lots of users is more satisfying (to me) than building things that nobody is using.
I was exactly this way ... maybe a year or so ago.
But honestly I've stopped being excited to type out code in my personal projects anymore either. I've become much more excited by what I can accomplish on my own in a small number of hours squeezed between work and family. I still experience this as a loss, but I'm no longer so sad about it, and moreso feel invigorated by the possibilities and opportunities that have been opened up.
A way that I have come to think about this is: I used to always be curious about the product management role. How exciting to come up with ideas and validate them with users. But I always demurred because it would be so frustrating to have to rely on other people to bring those ideas to fruition! On balance, I always preferred being the one executing ideas to being the one generating and validating them. But now I can properly do both things! (In my hobby time, that is, at work we still have this idea/execution split, at least for the time being.)
FWIW, I was just like you but then completely gave in and found enjoyment in the act of simply ideating and shipping. The gap now between idea and implementation is so small. At first I was depressed but now I'm in the acceptance phase of grief. We aren't going back, for better or for worse.
What we're seeing now reminds me of that pub dialogue about running in Back to the Future 3, paraphrased:
> Jeb: "If everybody's got one of these auto-whatsits, does anybody code anymore?"
> Doc Brown: "Of course we code. But for recreation. For fun."
> Jeb: "Code for fun? What the hell kind of fun is that?"
Heh, my employer kept pushing us to use Copilot. And over the last months the cli has actually gotten halfway decent... So I did start using it. Albeit sparingly because the token allotment was always pretty low.
Then they announced that they removed the limit/making further request just cost extra for them. That's when I started using it as I did for my personal projects I pay subscriptions for...
Then Copilot increased their pricing. Announced in April I think? But took effect this month. This Monday they announced that the limits are back in effect. So I guess I'll be going back to hand coding next week, as my tokens are about to run out ಥ ‿ ಥ
Corporate is always so silly. I mean I know how it happens: everyone just wants to get their bonus, so different management roles try to coerce the employees to do whatever best serves their bottomline - rarely related to whatever is good for the corporation... But it's always silly to live through it.
If I'm forced to do more work I don't enjoy, I expect to be compensated better. So now you have the cost of the AI and salary increase to deal with.
I think this is a larger question we should be asking. YES, we can build this world: a world where robots do our chores, serve our coffee, check us out at the grocery store, and let the AI agents do the parts of our jobs we love.
But SHOULD we? With great power comes great responsibility - and I'm getting the impression we're (quickly) building a world that isn't very fun to live in. We technically have a choice here - DO we want bots writing our prose and responding to our customer service inquiries?
I remember a book I read as a pre-teen, 40 or so years ago, about a kid who wanted to be "perfect". A wear a tree of broccoli on a string around your neck to learn how to overcome embarrassment. A perfect person never makes mistakes, and the best way to not make mistakes is to not do anything. Similar "requirements" of perfection and their expression are presented. The kid eventually finds himself in an empty room, by himself, doing nothing, wearing broccoli. Perfection was achieved, but at the cost of an extremely boring life.
I'm of the opinion that "true art"/cultural artifacts can't be automated by definition, as they derive their value from the human experience embedded in them by their creators.
OTOH, I think we absolutely SHOULD automate necessary "drudgery" type work wherever we can, but we're going to need a radical reconceptualization of how we distribute the spoils of economic productivity as a result. Unfortunately, I think the type of reconceptualization we'd need would entail a complete overhaul of many long-established and deeply-internalized concepts (rights and duties of ownership of intellectual and private property, decoupling of identity and occupation, etc.), and from everything I've seen, that will be a long and painful process assuming it's even achievable. (Especially in the US, where decades of pro-business messaging has yielded a culture that equates income-earning ability/entrepreneurial success with individual human worth. I really struggle to imagine a path toward unwinding that, but there's little chance it'll be a smooth ride.)
> customer-service employees who enjoy building relationships but are increasingly expected to supervise AI agents instead.
It sucks for the employees, otoh it might be the only way we're going to beat Baumol's Cost Disease.
In the past few decades productivity has exploded, but service employees have largely failed to increase productivity in any way because it's harder to automate these tasks.
It's the reason the costs of things like education and healthcare are downright extortionate, the reason you're paying back your college well into your fifties, the reason you don't call an ambulance for someone in the US because you don't want to ruin their life financially.
We may have to trade the personal fulfillment in these jobs for the broader affordable access to these services.
Education and healthcare are both ridiculously overpriced in the US for reasons that have little to do with service costs. Questionable financial systems behind these services are much more to blame.
>It's the reason the costs of things like education and healthcare are downright extortionate, the reason you're paying back your college well into your fifties, the reason you don't call an ambulance for someone in the US because you don't want to ruin their life financially.
You might wanna think again on that line of reasoning, because plenty of other countries have the same dynamics with respect to service employees, but they don't suffer the very US-only problem of ridiculous education and healthcare costs where calling an ambulance can ruin someones life.
My point is simply this: a person who helps individual customers in any industry isn't much more productive today than they were fifty years ago.
That may change, and it may benefit everyone except the people who get fulfillment in their life from the one-on-one human interaction they get from people who need services.
“If the government pays for it, it’s free!”
Exactly. You get it.
As a former first responder, I'm interested in hearing more about how AI-powered ambulance services would work. (related question: will the 911 dispatchers be AI?)
I don't think first responders are ever going to be at risk.
Administrators, on the other hand, are a massive part of the costs in the health sector (IIRC the Obama administration chickened out on truly reforming healthcare exactly because the number of administrators that would be made redundant would tank the economy). A significant amount of administrative work can be automated.
As a person who drives on a daily basis, I want to know why we don't have AI controlled stop lights and overall traffic control.
> It sucks for the employees, otoh it might be the only way we're going to beat Baumol's Cost Disease. In the past few decades productivity has exploded, but service employees have largely failed to increase productivity in any way because it's harder to automate these tasks. It's the reason the costs of things like education and healthcare are downright extortionate, the reason you're paying back your college well into your fifties, the reason you don't call an ambulance for someone in the US because you don't want to ruin their life financially.
Weird. I thought it was the fact that you have a cohort of people who are grossly overpaid to represent people who do none of the work yet expect an ever-increasing amount of value created by the work to be shifted to them every 90 days, no matter what, forever.
> We may have to trade the personal fulfillment in these jobs for the broader affordable access to these services.
Then you'll run into two problems:
1) no one will want to do necessary jobs without increased compensation, which is at the root of your analysis of "Baumol's Cost Disease"
2) at least in the US, you'll have a bunch of increasingly miserable people living in a society that gives them less and less to lose while increasing the availability of things that allow them to take out their frustrations upon themselves (substances) or others (weapons)
Yeah I definitely agree with this. AI has become very useful to me, but it has also definitely automated some of what used to be my favorite parts of my work.
I am having some success in working to acquire a taste for different parts of the work. But I suspect that this won't be an option for most people.
As always... no AI-hypester ever talks about Amdahls law (https://en.wikipedia.org/wiki/Amdahl%27s_law).
Most people don’t have jobs they enjoy, programmers have somehow escaped this; better pay and more “protected” time instead of doing what the rest of us were doing all along
They did till 2025. Now more than half don't enjoy it as the technical puzzles/plumbing is gone. The outcome oriented folks find it fun (and amplified up).
> That's going to give back the productivity gain and more, while also decreasing my quality of life.
but did we increase our EBITDA for the quarter?
> What's a 20% productivity gain
Where did the 20% number come from? I’d argue it’s way more than that (or variable, i.e. dependent on who’s using it/how it’s being used/what it’s being used on).
Having said that, the number, to me, doesn’t even matter. You could replace that with 200%, and it’d be just as true.
The vast majority of jobs are not full-filling or enjoyable. Because there were way more job seekers than jobs.
Programming was one of the ones which was, because there were fewer programmers than openings. Now that's flipping, thus naturally, the enjoyment is going to be sucked out of it.
The cynic in me has learned one is measurable and can go on a slide deck, the other is vague and hard to measure.
anecdotal but a good number of people have clearly stopped working as much or quiet quit. some even left the company or the industry. CEOs are also telling everyone to 5x their output, AI will replace them.
Whatever productivity gains models are giving us is being eaten away by other factors.
Don't worry. They'll find some freak that actually enjoys it and is even willing to be paid less!
It's like if your career switched from solving puzzles to filling out TPS reports.
For me it feels less like filling out reports, and more like mentoring an intern who can search for stuff really quickly but forgets everything at the end of the day due to anterograde amnesia.
Except the intern is trapped inside an iron lung and must communicate entirely by text. And also has zero real creativity or self-motivation.
> What's a 20% productivity gain if I constantly feel deflated by work that used to energize me?
I think it's called 'capitalism'
> What's a 20% productivity gain if I constantly feel deflated by work that used to energize me? That's going to give back the productivity gain and more, while also decreasing my quality of life.
This has been the story of humanity since the industrial revolution.
I mean that sucks, but I have many tales of people, who were passionate and outstanding in what they did, and were rewarded with a leadership position for their efforts.
Now they get to fill out excel sheets, babysit people and sit in planning meetings.
Then they'll fire you, find someone who will work for the wage that the now-degraded productivity justifies, and get them to use the AI.
Many successful people in their 40s could quit work already if they only decreased their monthly costs.
I think it is more important than ever to manage your wealth in a way that sustains you from capital alone in a world where employment gets progressively more toxic.
The way to achieve it is buying maintenance efficient and cheap car, make renovations smart, make good choices all around to minimise expenses. Operate your life like a corporation. First, cut the expenditures.
For me the ability to do whatever I am interested in at the moment, is worth almost any sacrifice.
Then I can seek one time contracts or short time jobs that fulfil my mental needs.
My monthly expenses are no more than 5000 dollars and mostly consistently less than that.
Which is okay because the money spending doesn’t bring me any joy nowadays. The money isn’t what gives me happiness. Only other people can provide that and activities that are dirt cheap usually, like reading or broadly understood hacking
Consumption is a short lived and deceptive joy that causes more guilt than whatever dopamine it is worth really. Governments hate people like me which means I must be doing something right.
I think that's a very simplistic take, it doesn't account for people who need to support children's education and/or healthcare for parents.
Both are areas where costs are going up at a rate which is much higher than official inflation rates, and comp increases often don't even match inflation!
So the need for working till you're forced to give up still remains.
Also disregard this because I don't live in USA with its ridiculous healthcare costs even if my conversion of numbers accounted for purchasing power parity, it did not calculate monthly healthcare expenses that in my locale are only 1/7 of median national wage per month flat which I assume is a much better ratio than in USA and very predictable.
As per children, contraception exists and you should use it as kid is a dubious and too risky investment. You could for instance raise a little sociopath or two.
Healthcare for parents is for me at least letting them into a boat in the ocean when they get first alzheimer's symptoms like in Hemingway Old man and the sea. Native Americans used to do it. They just released their old guys to the forest or steppes. Which reinforces my point about the kids being rather bad investment. Do not count on that.
It's objectively more wise to save 500k-1mil from raising a kid to invest that money and spend it when old, hiring a professional caretaker. Have friends, spouse, not kids.
In any case it probably only matters that USA healthcare costs are so exorbitant. This is the crux to why it could never work for americans unless you are at much higher financial level than any sort of frugal upper middle class specialist in 40s.
It's the only western country where retiring earlier is so expensive. I thought maybe it would make up in the extra money you can make and save in taxes without VAT but no, after research I think that healthcare is so much pricier that it is incomparable.
So it is not simplistic but it's different environment altogether, one in which it actually works and I am really grateful for that.
Most people don't have the luxury of finding joy and meaning in their work. You aren't hired to have fun, you're hired to create value and wealth for your employer. Just do what literally everyone else does and grind through it until you get a pension and hope it's enough to let you die with a bit of dignity.
Here in stupid town we stab ourselves with knives. It hurts and we get infections but everybody does it. We don't even remember why we do it any more but we do so you have to as well.
The real problem is the amount of value that gets left on the table.
Also, mental health is just as much a part of health and well-being as the physical.
We know why we do it. We do it because we live under a system that requires most of us to bleed for our supper. And now tech is being normalized to work like every other job, and techies get to feel the knife. A lot of people are going to be shocked to realize what their actual relationship to capital has always been.
If we know, why would a lot of us be shocked? Seems contradictory.
That's probably too much a troll so while I get that you are grinding your anti-capitalism axe, you're also seeding hopelessness and trapped falsely fatalistic thinking. It has gone this way but it doesn't have to and the whole arrangement is a local optimum with much higher global optima available.
In fact there are many of us who have created a parallel life and extend ladders all the time (within our capacity) for others to join us.
That is definitely not what everybody else does. Some people try harder to find the things that they enjoy doing that provide value to others.
> That's going to give back the productivity gain and more, while also decreasing my quality of life.
Unfortunately though, what does that matter? Your employer does not care how do you feel. You are paid to bring them benefits, they aren't running a charity. If you do feel down, that is your issue and you shouldn't let that influence how you work.
Just to be clear, I don't like that either. But it is what it is.
Mental health is health. Negative mental health impacts are particularly significant when your job is to perform mental labor.
It's not defiance, it's reality: if you drag down my motivation, that's going to drag down my productivity. I don't really have much active control over that.
But that's the point, employer doesn't care why you're sick, only that you are. You create "problems", you will get replaced.
That is when you unionize and make them do what you want.
> That's going to give back the productivity gain and more, while also decreasing my quality of life.
Is everyone entitled to a high quality of life?
If not, then who draws the line as to who deserves what benefit in life? You?
This poster wasn’t claiming they were entitled. What kind of question is this?
> Is everyone entitled to a high quality of life
Yes.
> Is everyone entitled to a high quality of life
Only finite land available in high QoL areas such as Orange County California, New York City, or Hawaii, depending on your lifestyle.
You want to tell us who is allowed to live there?
Because sure as hell won’t fit all 345 million people in America with a desire for higher QoL
One imagines that quality of life ought to increase as technology evolves and the economy grows.
People are definitely entitled to complaining about decreasing quality of life and not liking causes of such changes.
Driving everyone's QoL to be as bad as possible will lead to increasing enshittiffication in the entire market.
Consumers will be spoiled for choice between deeply mediocre options.
Besides, what's the point of adopting new technologies if it's not to increase the quality of life? If everyone just exists in service of the product development lifecycle, who and what are the products actually for?
In my experience, for every customer support agent that really wants to help people and cares about their problems, there are at least ten who don't even read what you wrote and answer with prefabricated blocks of text that have little to do with what you asked. If AI customer support actually tries to understand what I ask of it and help me, and there are still (motivated) humans available for the more tricky cases that AI can't handle, that might be a win...
> If everyone just exists in service of the product development lifecycle, who and what are the products actually for?
Anyone holding passive index ETFs in their brokerage / 401k / pension accounts.
6 hours a week is low, unless its the average spread across industries. I think I spend more time in Claude Code via the CLI versus any other app I have on my laptop.
Like others said, the frustration is when it gets something so wrong you just think "wow, how'd you mess that up?" but when it gets it right its kind of nice. I also dont like that I basically tell Claude what to do, and then either go to busy work or waste time on the internet.
I kind of enjoy exploring black boxes, trying how different inputs are mapping to differences in outputs. It's kind of like hacking. The problem is, they keep altering the box.
The box is stochastic by design, and has an untraceable amount of complexity between its context and output by nature.
It may be fun to look at inputs and outputs, but it's not hackable and trying to map one into the other is more like astrology than a science.
It's copromancy. Picking through the clanker's doings in an attempt to predict the future.
Thanks, you taught me a new word today! https://en.wikipedia.org/wiki/Scatomancy
It feels like Greek mythology should have some metaphor for "apparently simple structure that is so complex it leads anybody that studies it into madness". But I can't think of any name to put there.
Maybe the idea of complexity is too modern.
No but you see, I have a system! /s
(I spent too long by the horse racing track)
The third sentence got to what my objection was going to be. It's fun trying to make the thing do what you want it to do! That's why many of us like computers. It's the randomness that sucks and makes the process unsatisfying.
That's just just a slot machine
Working with AI is trying to reduce the probability it'll pick undesirable paths. It's an exercise in trying to avoid what you DON'T want.
I suppose it's the same as asking someone else to take care of a feature and hoping they understand what you have in mind. The difference is that there's a lot of context that's shared between you and a human developer that is simply absent with AI.
I wonder how many hours of old stye non-overhead that is. And by overhead I mean mostly meetings. Then again maybe generating paperwork is included so that eats bit into it.
An hour out of every work day doesn't seem low to me.
Welcome to the slot machines!
I theorise that many social ills come from workers having less pride in their skills and achievements, and a greater sense of social alienation, due to automation.
If you spend countless hours at work, and you partially define yourself by your work, and you realise you are easily replaceable then I cannot imagine this comes without mass social malaise that manifests itself elsewhere.
When you know you're essentially babysitting the workhorse to ensure it doesn't go off the rails, I can't see job satisfaction, and the social consequences of such, increasing.
> When you know you're essentially babysitting the workhorse to ensure it doesn't go off the rails
I don't have data to support this (other than, I guess, my LinkedIn feed), but my impression is that the management class is pushing AI _way_ harder than the worker / craftsperson class.
And if that's true, I think it's perhaps because it's something they understand: you tell AI to do something, and (with varying degrees of success and less complaining)... it does that thing.
To the extent that I've seen craftspeople adopt AI, it's been because they recognize its usefulness as a tool to further their craft. I don't meet many craftspeople that enjoy watching any[one|thing] do their work for them.
I agree with this. The stack seems to become:
LLM worker > Harness, Agents w/ skills > Human oversight/input
This is similar in structure to many teams I work with, something like:
Dev/SE/etc > PO > Manager/Director
Or whatever your current org structure relates. The LLM worker and Harness/Agents compact down to one human layer.
Now with MOE LLMs, the LLM layer is breaking into like:
LLM worker > LLM router > Harness, Agents w/ skills > Human oversight/input
Does that mean the Human element can be condensed to a single Manager with the right skills? A Director above them? Is the VP above them directing the agents?
Is this another variation of Conway's law, where orgs design systems that copy their own communication structure? Seems like that is how my Manager/Director approach it. Then again, they are making slide shows and obviously AI assisted reports, not something that needs to be stable and responsive for the entire product lifetime.
But to your point, the manager sees it as a structure to manage to increase productivity. The craftsman sees it as a tool to further their craft. Each is driven by a different methodology and use case. Can that mesh unless specifically directed to throw AI at no joy work?
> I theorise that many social ills come from workers having less pride in their skills and achievements, and a greater sense of social alienation, due to automation.
Welcome comrade. https://en.wikipedia.org/wiki/Marx%27s_theory_of_alienation
I agree entirely. Even in an idealistic fully egalitarian post-scarcity society, to truly be happy I think most people would need to do work that they can feel a sense of accomplishment about. The problem is that work at most jobs is increasingly just toil. Any possibility to scrape some tiny flakes of satisfaction out of the toil gets removed, often for no good reason.
I'll be in my woodshop!
There's been a fair amount of research done on this, and while Marx's alienation theory isn't totally out there, what makes the most difference for workers is autonomy and purpose. We need to feel like we have some control over what we do, and like what we do is meaningful.
That's easier with things like farming, but it's totally possible with highly-automated jobs; you just can't have managers treating us like we're machines.
Are you a communist?
Follow up question: Are you aware of it?
Believing pride in ones work is important, a communist makes you not. And sweeping ideas away by categorising the idea is the destruction of debate.
Sorry, I was being provocative more than anything. I should've said nothing.
I see some similarity to how I felt when library management/wrangling became a huge part of software development.
In the last century I enjoyed crafting my own 'libraries' of functions that I could then use on the projects I worked on. As time went on, there was less and less of a point doing that as the odds rose near to 100% that there was 'a library for that' thing I was working on, so I was encouraged/forced to download it and use it.
It solved problems and was quicker than writing bespoke code (and libraries were hardly a new idea), so the logic was hard to deny, but I enjoyed my job less over time. Now I've risen up the ranks and now code mostly for fun (yes, I use AI to write functions for me) I look at what it must be like to enter the industry and think it all looks very different to how it did when I started.
You could argue that AI has done this much faster than it did in my early career, so people have less of the 'boiling frog' experience I had, and more of a 'sudden shock' to the system.
It's sad, but I've been doing this to other industries all my career, so I can hardly complain.
You pay per token, even on subscription models the limit is tokens.
If I was valued at 1 trillion dollars, and I was in the hole enough to sink a couple small countries' GDP, maybe I would slowly start to optimize to maximize token usage.
I want to sell tokens, how do I sell more tokens? Not by doing the same work in less tokens, that's for sure.
This is like if you pay me by the hour and then excitedly tell me that you keep paying 10k a month and it's great. I will most certainly not work faster, in this hypothetical, if you tell me you love spending money because it gives you a dopamine rush. I would probably spend a couple more hours REALLY thinking about the task, maybe writing some docs nobody will read, maybe considering multiple options, doing benchmarks, doing research, and then later maybe ill do the actual task as well.
Im not saying these AI companies are scamming us, but the incentives are there and extremely clear. The only thing currently holding it back is that there is some vague kind of competition.
Anthropic just said they're going to limit the ability of the new model to assist with building foundational models.
The ability to assess the user's goal and tune the inference accordingly is there. So, what methods are there to tell if a provider has their thumb on the scale?
Plodding along a longer path to the same goal would burn more tokens without necessarily a decrease in solution quality. Maybe a few more docs.
It'd probably still be worth it to make tokens cheaper. Jevons paradox [1] seems to apply here, where making compute cheaper unlocks use cases that were previously priced out, increasing overall demand.
[1] https://en.wikipedia.org/wiki/Jevons_paradox
I spend at least 6 hours a week arguing with bots owned by other teams, as I’m unable to reach a human before I bypass their bot. 10k person company, clients are paying for my time.
Soon all communication will start with bots talking to bots trying to find a human. It's already happening with phone calls.
I would be tempted to send my own bot to do that drudgery
It may be that they’re protecting their time.
Right. Somewhere there’s a dashboard which lists those 6 hours as time saved.
They are, eng makes 600k/yr and I make 2, so their times is worth triple mine. It’s cost effective to make me toil and drown in frustration if it reduces ticket volume.
Just build a bot to bypass their bot.
Corpo bullshittery is the best kind of work. Get paid without actually ever doing anything. Its heaven.
Being alienated from the outcome of your labor is far from my idea of heaven.
Not if you enjoy making things and take pride in your work.
Actually having to work and take responsibility is extremely exhausting compared to corp life, I experienced both, I currently have no corp job, am responsible for a lot of apps and hate every part of it. Boreout is so much better than burnout and I never worked a job that allowed me to have none of them.
That's some odd image of heaven.
I've found that setting good guardrails, and running in a sandbox so that the agent doesn't keep asking tedious permission questions, makes things go a LOT smoother.
Generally, I spend anywhere between 15 mins and an hour setting things up (depending on how well the project is set up for AI work), and then set the agent going, coming back in a half-hour to an hour to check its progress. Generally, the tooling keeps it honest (for golang, forbidigo is AWESOME). 80% of the questions the agent asks me require a lot of thought. 20% of what it does needs correction.
The other thing to remember with LLMs is that they are NOT human, and won't react in a human way. So you'll see strikes of "brilliance" followed by the absolutely bizarre. But good guardrails keep that to a minimum.
> sandbox so that the agent doesn't keep asking tedious permission questions
> 80% of the questions the agent asks me require a lot of thought. 20% of what it does needs correction.
I've found even the permissions questions give me veto power over fruitless lines of exploration, especially in planning mode. For instance, it wants to use tools I don't have installed to access information that I have made available elsewhere? I get a chance to override this decision by declining the permissions check and redirecting it. Feels tedious, but helps me understand what information sources are influencing it. I head off a lot of bugs this way.
I never let it go into planning mode, other than to output a plan file that I can audit before giving it the go-ahead to implement. After that I don't want to be bothered, so --dangerously-skip-permissions keeps all but real questions out of the loop, and I can do something else while it works rather than babysit.
This is my impression too. Whenever it needs permissions outside of a small set of defaults I've allowed, it's often because it's trying to do something ridiculous that it doesn't need to do.
I think the yoloist counter-argument is "So what? Let it. It'll take longer that way and consume more tokens, but you can work on something else in parallel instead of being hooked in to this one session".
How often are you going into new projects and spending up to an hour on set up? I'm really just asking to get a sense of what "Generally" means here.
I do it with every project I go into. First step is setting up the documentation so that the agent can navigate it quickly, knows the idioms, knows the test gating procedure, design principles, coding standards, testing policies, etc.
Once that's set up, I spend time laying out the planning for whatever feature or fix is being worked on. For fixes the agent is pretty quick and usually needs little guidance. For new features it's best to have more of a hand on the tiller.
It doesn't change the premise.
AI should be assisting us, instead it's doing the job and it's us being an assistant to it. This is a monumental shift that people seem to be missing in how knowledge working is changing and it's going beyond mere coding.
Guardrails, prompts, whatever, it's us helping it doing the job, not the other way around.
Opus 4.6 was the last genuinely good assistant LLM, but since then it's quite clear that the training/reinforcement is focused "given prompt -> do task" so it's behavior is more and more about doing it itself, not helping you. If you try to use it as an assistant it just sucks and is perma wired into finding the solution. Many times I want it to help me investigate, and his answer will still be focused on the fix, not answering my questions.
4.7 first, 4.8 later and fable are absolute disasters as assistants.
Fable in particular is so "intelligent" that it will push with very strong and intelligent takes even if it is completely wrong.
I have never disliked our job more.
Wow... Our experiences have been very different, then. I've found each upgrade of Opus to be a noticeable improvement in its complex reasoning and delegation capabilities over its predecessor.
To me, this feels in many ways like a technical manager or team lead's job, where I guide the process along using my knowledge and experience, and then let the agent fill in the rest (to the best of its ability).
The agent can't really learn from its mistakes (at least, not without consuming precious context), so I apply a blameless postmortem process, updating the guardrails whenever it goes astray in the same way more than once.
And really, I'd rather be contemplating the more difficult and interesting questions of architecture, environment, ergonomics and market fit, so it suits me fine.
Same here. The power upgrade going to Fable in particular is quite impressive.
> Wow... Our experiences have been very different, then. I've found each upgrade of Opus to be a noticeable improvement in its complex reasoning and delegation capabilities over its predecessor.
I haven't stated that it's not more capable nor more "intelligent", it's the opposite.
I will try to expand on what I mean.
LLMs "character/persona/tendencies" are increasingly less about acting as an assistant and more about finding the solution itself.
I use AI in a specific way: he assists, investigates and answers my question. I do the coding. It is increasingly difficult to use it as such, because it quickly jumps into giving me solutions instead of answering my specific questions.
I'll give you few examples.
I asked it to investigate DNS handling details in phoenix emailer module work, he did very little investigation and jumped into why I should've used magic links instead. Instead of assisting me in my research, it was hard wired to solve the problem (the wrong one, with a very wrong solution).
Today at work, I had a problem with batching, I wanted to understand if batching was even needed at all for our use case, and he kept circling around how to fix the batching bug instead. That's not what I asked it to do, yet, it jumped to the "solution".
I am increasingly frustrated by these models "personality" and tendencies that are unhelpful to assist me doing the task at hand and more on it doing it and me merely assisting/supervising.
Sure, very detailed prompting on how he has to act helps, but wait few turns and he drifts again to his default solution vomiting state.
Which makes me think that these models are hard wired on this mode of operation by consistent training and reinforcement of jumping from prompt to code solution.
Ah yes, the agents by default are very "implementation" oriented, which is why I instruct mine to never implement something without formulating a plan first for me to approve.
Another thing they tend to do is rely on their own context -> memories -> training data. And if that's wrong then they'll continue with it until you instruct them to research, after which they usually get the right answer.
I've noticed that the newer models keep track of what you type so as to anticipate what you're likely to say. For example, today Opus 4.8 said "You usually don't want me to commit until you've checked, so the change remains uncommitted."
I think this is just a misunderstanding of how most technology has always worked?
Consider what is happening in most construction sites. The heavy work is absolutely from the technology on site. But without people there to oversee it and keep it working, it would fail.
And that is almost certainly true at any industrial site. Indeed, look up videos of high tech looms. A large portion of the technology added to them are so that the operators can locate the fault and fix it.
The problem (okay, one of the problems) with renting other people's models is, as you mentioned, that they can and will change out the model without notifying you ahead of time, and you don't always get to control which model you use. (They might decide to retire it, and you won't be able to get it back if they do).
Which is why (well, part of why) I think the long-term trend will be towards self-hosting models. Right now the frontier models are far enough ahead of the self-hosted ones that there are lots of people willing to pay by the token to rent someone else's model, because they get more value for money from that than from self-hosting models.
But the frontier companies won't be able to keep up their current levels of expenditure forever. At some point the investors are going to say "Hey, so, um, when am I going to see some return on my investment?" and then the current subsidized subscriptions (including the one my employer uses) are going to go away, much like what happened with Copilot this month.
And then the locally-hosted models are going to suddenly look like a more attractive picture. Because where you might have been willing to spend $100/month/employee to rent time on models in someone else's data center, you might suddenly balk at spending $500/month/employee. You might say "Hey, you know what? A $50,000 up-front capital investment is only, what, one month's worth of subscriptions for our 100 employees? Yeah, okay, I'll approve the hardware purchase. Get that self-hosted model set up and then we'll cancel the subscription and switch over."
Not everyone is going to do that. But once the locally-hosted models are good enough, the first few people who do so and report success are going to start a snowball effect. And it will likely be driven by money first, but it will also have the effect, that people will slowly discover, of meaning that you can better predict the model you're using. It will continue to work the same way next year that it is working this year; or if it doesn't, it's because you chose to install the new version.
And when that happens (I'm saying "when", not "if" because although it might take some time, I think it's inevitable in the long run), the frontier-model rental companies are going to struggle to stay afloat. Except for the ones who saw this coming and transitioned to a non-subscription income source somehow (maybe by selling licenses to self-host their frontier models for $$BIGNUM), or who have some other revenue stream besides renting out models.
AI should be assisting us, instead it's doing the job and it's us being an assistant to it.
If you're a manager and you ask a report to do something and they come back with a question, does that mean you're now their assistant?
I give agents the tasks, I answer their questions, I make choices about the tradeoffs in their plan, I supervise their implementation, I review their output, I have them walk me through things. In what way is this not delegating to them and managing their work, just like a more junior employee?
That sounds weirdly gendered even though there's no reason it should be.
Are you getting LLMsplained? :)
Well... as a human software engineer, I've been the one with very strong, intelligent, completely wrong takes. The question is, are the LLMs improving faster than you can improve a junior dev? And is their ceiling as high?
Your experience pretty much mirrors my own. I hate to be the 'they're holding it wrong' guy but there's certainly a lot of people out there that have no real idea how to effectively leverage AI.
That’s a problem with the tool not the people. AI is marketed literally as writing one sentence and having some app perfectly output. Just check any of the landing pages for Claude code or codex or GitHub copilot…
No, it literally isn't. I just looked at the landing pages for Claude Code, Codex, Cursor, and Copilot, and literally none of them have anything about "writing one sentence and having some app perfectly output", or anything remotely like that. In fact, just the opposite: they all make clear that they're built for ongoing collaboration with AI, and have detailed descriptions of what that looks like. No one advertising the idea that you can one-shot perfect apps with these tools.
Right from the codex homepage in the first screenshot “Create a compelling launch hero for the new Codex app in openai.com/codex”
Talk about moving the goalposts!
i've seen a number of articles claiming things like "devs self report they'er +x% more productive with AI, but actually they're -y% LESS efficient!". and i think that this is explanation for why.
as a boss (or researcher) i'm going to measure productivity based on amount of output per hour that i'm paying you; as a workers, i'm going to measure productivity based on amount of output relative to the amount of effort i'm putting in.
so what may be happening is that bosses see that output is at 80% (productivity down!) but workers see that they can give that 80% output with 40% effort (productivity up!).
Not sure among devs, but I do know that in other positions in typical corporate bureaucracy, people have a propensity to not report their own automations or productivity gains upward, because the reward structure isn't there.
Early on in my days as a sysadmin, I automated a ton of my role when the rest of the team was still doing ClickOps. The reward for doing so was more work and expectations without the additional pay increase to justify my new found productivity. That happens all over the workforce, and so people will just keep it to themselves. I learned my lesson at that first job real fast that if I'm able to have the same, or greater output, for half the time, I keep that to myself so I can use the automation to free up my own time instead of have it filled by the company.
I wonder how much of that is happening now with AI in non-technical roles.
https://www.youtube.com/watch?v=OwfNjGxa_D4
> so what may be happening is that bosses see that output is at 80% (productivity down!)
If an initiative produces only 80% of the previous results and you’re paying large token bills on top of the same wages, the AI is going to get cut off.
> i've seen a number of articles claiming things like "devs self report they'er +x% more productive with AI, but actually they're -y% LESS efficient!".
Are you thinking of the old METR evals? Their more recent evals showed an actual performance improvement.
The old report is still circulated as bait for AI skeptics.
I think the old report you're referencing is this [1] from July 2025, but I can't find a new report. This [2] links to a new dataset at the bottom (that maybe shows improvements?) but it seems like they chose not to write it up because of perceived flaws in their study. Is there a more relevant report I'm missing?
[1]: https://metr.org/blog/2025-07-10-early-2025-ai-experienced-o...
[2]: https://metr.org/blog/2026-02-24-uplift-update/#wider-adopti...
I read this today and found it super valuable in evaluating METRs research.
https://arachnemag.substack.com/p/the-metr-graph-is-hot-garb...
> so what may be happening is that bosses see that output is at 80% (productivity down!) but workers see that they can give that 80% output with 40% effort (productivity up!).
So why is it that the bosses are the ones that are so enthusiastic about adoption?
Well the obvious is that, monkey-see-monkey-do, and they don't want to miss out. But the insidious externality is that they are the most likely ones (and higher ups) to be invested in nvidia and others, so when they push you to use, it creates a viscous cycle.
Do you have any studies in the last 6 months showing performance decreases?
March 2026:
https://www.faros.ai/blog/ai-acceleration-whiplash-takeaways
Faros would tell you shops are shipping 16% more PRs with heavy AI use.
https://unessays.substack.com/p/talk-is-cheap
I think that's wrong because it doesn't adequately account for the quality and rework their data show.
It's horrifying how many people I see here being threatened into using AI. I guess I'm fortunate in that I have a choice.
I've used it plenty, more recently than ever before, but I'm coming more and more to the conclusion I don't want anything to do with it.
Botsitting is low-skilled. Low-mental reward work will become low-financial reward work before long.
I was initially worried that I had to keep up, learn the new tools or get left behind. But I'm beginning to see it as an entirely different domain of work, and one I'm fine not doing.
I think I'm best off preserving my skills and autonomy, rather than fighting with an agent and fretting about tokens for the foreseeable. I've no desire to be a botsitter.
Worst case scenario, this really is the future, and I become unemployable in tech: I've done other sorts of work in the past, things I'm happy returning to if this is the future.
I've a hunch that's still going to be a while off. I hope so anyway.
i was a senior designer that basically had to reclass to a manager/senior manager role just to keep a job that didn't touch AI as much.
the wild thing is, i liked my job 3 years ago. i liked the satisfaction of creating for sales. they could be annoying, but it was minor normal (human) corporate squabbles. now, when they're explicitly bragging about getting rid of designers and treating us as sub-human, it's creating resentment and misery. personally job satisfaction sharply decreased in the last 6 months let alone the last 3 years - i'd say from 80% happy to 20%.
personally, i think suicides are probably going to be on the rise in a big way specifically due to this; and that's just the people who can hang onto a job right now; cannot tell you how many designers and animators are out of jobs and desperate.
My favourite personal experience is how they disabled yolo mode in Claude Code at my workplace
My challenge has been trying to manage my higher-level context. I've gotten a pretty good setup where I have project-level orchestrator agents that can spin up workers to implement tasks with minimal oversight, and the resulting work is usually quite good (especially after I give it the mandatory "make the comments less verbose" refining, etc.). But that means I'm doing even more context-switching. I've gotten to the point where I have a half-dozen draft PRs that just need my review before I tag my colleagues, and trying to dig up the context from all of those tasks can be paralyzing.
As always... no AI-hypester ever talks about Amdahls law (https://en.wikipedia.org/wiki/Amdahl%27s_law).
There was one good analysis from that viewpoint recently: https://x.com/addyosmani/status/2059844244907696186
Bot-sitting is the new long compilation times.
This is something I am very curious to learn from others since I am, relatively speaking, quite new to HN.
HN posts seem to go from very supportive to very critical of AI with very few articles striking a balance.
Was there a time or technology phase in the past (I'm thinking perhaps when everyone was going over to GraphQL just because of the hype) that you recall being similar to this? If so, how did you maintain curiosity without feeling yourself go through the extremes of "This is the worst / best thing to ever happen"?
HN has a huge user base. If the article is pro-AI, you'll see top comments praising AI. If the article is anti-AI, you'll see top comments deriding AI. The article and headline is just an attractor to those whose opinion is aligned. It's not unique to AI, basically every thread has your typical nerd-fights, but AI has been particularly polarizing for many reasons.
True but it also depends on the initial audience that sets the tone, and the time zone. In my empirical experience, pro-AI posts fare much better when West Coast is awake, and response is more muted when Europe is awake.
We're also forced to use it. I can't stand it. Want to just quit the business altogether. Maybe become an electrician or something.
I just started using Claude Code for my work as a sysadmin. For my work, it's great. I don't need to wrestle with MySQL joins, claude gets even the most complex ones right WAY faster than I would. Same with new Terraform stuff. Things that would have taken me a day are cut to less than an hour.
So for my work, it's made me much better at my job. Much faster and more accurate.
I don’t know.
I can write a simple query before Claude finishes reading, querying the semantic layer, checking my files, then writes a query that I have to approve, reads the results, hides them (ctrl+o usually works), and gives me a summary.
We’ve reached this inflection point where it’s faster for me to do most tasks again.
I’m sure fast mode costing more money plays a role.
Personally I find LLMs absolutely terrible at writing Terraform and full of hallucinations. But also Terraform is my bread and butter and our use/workflow is fairly advanced. And we're multi-cloud + baremetal. That wasn't an area where I was going to get a ton of value out of LLMs anyway.
I don't see a lot of talk about how AI development breaks the old feedback loop of write code, watch it run, change it, repeat. I really hate sitting around waiting for the agent to get done planning, reading the plan, then waiting for the agent to get done coding. It's those 5-10 minute windows when its working that really sap my patience and suck all the fun out of our jobs. Writing code by hand is just more fun.
> Writing code by hand is just more fun.
This is something that I don't see discussed a lot in these conversations, but its true for a ton of folks.
I didn't end up with a career in tech because I wanted to tell a bot to do the fun part of my job for me, leaving me only with the boring tedious parts. I didn't sign up to be a full time code reviewer, and I certainly never wanted to be a manager, yet alone a manager of bots.
It also can't help but spark feelings of "Why am I getting paid 6 figures for this??" and that makes me nervous for the future.
I imagine the engineers and assemblers in factories pre-assembly line felt the same when things started getting automated there. There's an element of craftsmanship that gets taken away as the product moves from being artisanal, hand crafted to mass produced.
I wonder if its too late for me to pivot to hardware
Yeah its hard to deny just the raw throughput from the AI. Like it really is doing work in hours that would take me days.
But those times when I had to drop down into a repl and play around with the output of a method. Or try different ways of doing what anyone else would think is boring, like array manipulation - that's a lot of what I actually LIKE to do.
A big part of me just hopes I can hang in there for another... decade, or two. Then I can retire! Maybe.
> another... decade, or two
I’m rooting for two! :-)
I don't mind the workflow since I'll spawn new agent sessions in new terminal tabs until my attention is saturated by round-robin'ing through them.
It's actually kinda pleasant, especially when I consider all the tickets I'm not excited about doing. It's prob worth focusing on that aspect of it.
You can still write code by hand. Just do that, you will run into tasks that are too boring, those you can do with an LLM.
I think a lot of developers who enjoy writing code are in an internal battle knowing they could write it themselves or AI could do it and probably do at same speed or many times faster. It's a contradiction as a developer because we always want to do whatever it is we are doing at the optimal speeds but we also like writing code. So, you're being forced to choose where you stand at a fundamental level.
Absolutely. And I'm saying that the fix is to decide to write code. The AI isn't faster at producing good quality, well designed code that you'd write (if you're quite competent, which anyone can become with practice).
The AI code generation speed is a lie, it doesn't get the same work done in less time, most of the time it gives you a different version. Thinking through the problem yourself can yield better results for the overall product, in my experience.
I am actively developing on 4 projects fulltime. Kind of like double or triple dipping but just 1 employer. With a Claude Code license I'm running 2-4 CLI instances with AI. I'm botsitting 6-8 hours a day and getting more done then every. I love how much I can build using AI
This kind of reminds me of an article that I saw on HN ages back, there's like a subset of office workers who automated their Excel jobs, and just show up to work, read books, and do literally anything, while Excel does their work for them, and they collect their paycheck.
I don't know what they're complaining about. AI has freed us from the drudgery of craftsmanship, letting us focus on the important stuff—managerial and administrative work!
(There's a reason why I call it the MBA's stone. It transmutes all knowledge work into a problem of management.)
For me, AI can sometimes create a false sense of productivity. It's similar to how in the past, people would spend time creating the perfect setup with notion templates, pomodoro timers and productivity tools, or tweaking their environment for maximum productivity, instead of actually doing productive work.
But now it's happening at the company level: "We're going to add a chatbot to increase productivity! Now MCP tools! Then agentic workflows! We’ll add skills, and now productivity will go up! Maybe loops will do it?"
It is surprising! I would have thought it is at least 6 hours per day.
Understanding what is going on with AI productivity is … frustrating to say the least.
The best I can say is that genAI is a self reported a 20% efficiency boost, and for a very (very) small group of people, it’s maybe a 2-3x boost. (And if you are at a frontier lab, you go fly into the big bucket of exceptions)
At this point, for most use cases, AI productivity is either the equivalent of giving people 3D printers, and seeing little benefit, or signing up for an outsourcing service, just without the development of human capital anywhere.
>Understanding what is going on with AI productivity is … frustrating to say the least.
Agreed. I think one of the hardest things about it is that productivity != value. You can push all the code you want, but if it's not driving revenue up or cost down, it doesn't matter economically.
Here is the best data I've been able to find. An observational study of 4000 teams over 2 years across many different organizations. Data gathered from their task management, version control, and CI/CD tooling. Critically - this is not survey data. It's much more direct measurement.
https://www.faros.ai/blog/ai-acceleration-whiplash-takeaways
Faros argues that teams are seeing about a 16% throughput improvement (PR merge rate) with heavy AI use.
I argue here that their data actually indicates negative absolute impact on throughput.
https://unessays.substack.com/p/talk-is-cheap
I think it depends on how you measure the boost. If you are talking about generating a first draft then yes, the boost is there. If you’re talking about completing the project in all well tested and architected aspects, then overall there really isn’t a boost.
6 hours of debugging and docs reading is not equal to 6 hours of prompt fiddling. The return of value beyond the few fixes applied will be almost nil from the fiddling.
Isn’t this just the new type of work? Human in the loop of automated processes?
Welcome to the factory!
Like Chaplin in Modern Times, we will tighten screws until we lose our minds.
Yeah, Amazon warehouses are just the same. Humans are only used for tasks beyond the comprehension or physical ability of a machine at that point in time.
The problem is, we haven't had the debate on a societal level if we want to go the star trek route (aka, we give our darn best to automate everything so that humans have the time to do whatever they want) or the realcommunism route (we ward off automation so that we have jobs for people).
The result of that debate not having been made is the third possible outcome - rabid capitalism automates everything as soon as it is profitable and lays off the humans, focusing on getting higher margins out of less people if need be; the best example for that IMHO is Disneyland or Vegas going on ridiculous nickel-and-diming tours. In the end however, there will be no one left any more who has employment and we'll be in for quite the riots.
Don't worry, soon you won't have a job.
6 hours which could be spent playing Doom!
AI is so underwhelming. I can't believe we are spending millions on this junk.
'Botsitting' -- that word is going into my 2026 lexicon! :-)
“the incredible ground-level utility that many of us on HN celebrate every day through undeniable, massive productivity gains”
I’ve been told before.
I'm yet to be invited to the celebrations.
I could care less about bot sitting (haven’t we always written our own automation?), but it’s botsitting the unverified slop that people send you that fuels frustration. I thought I worked with competent people who respected me
Our product lead/manager recently sent me an AI generated PRD (complete with a Claude Code spec!) to build a core feature which we have had for over 2 years (and is a highly used core feature by our customers).
I just can't imagine tanking my trust with my coworkers by doing something like that.
So we're now in this world where everyone is instantly 10x more productive at turning their thoughts into code. Now, think about the coworkers you've had that are middling to mediocre. Do you want them to have a tool that makes them 10x more productive?
That's what I wonder about, what happens to all those folks.
In the worst case it’s an even greater “productivity” boost for those mediocre folks, as they will also undoubtedly be less critical of its output.
Maybe this is the AI layoff wave we'll see. Sorting out incompetent team members.
the ones who spend all day telling the bosses how great AI is?
It's not a lack of respect for you; it's a lack of respect for the work itself. That lack is being rewarded and encouraged.
Managers will be sure to tell you how much they respect you. Ask them if they respect the work and you'll get a blank stare.
Your coworkers haven't changed. What changed is that people can hand off work they never had to think through themselves. So you don't know what they checked and you don't know what you need to. You just have to read the whole thing.
*couldn’t care less
Guys you are privileged to do a job that you love. The privilege is even higher when the things you love aren’t even needed anymore. If this is the case you are on the verge of losing your job. Key word is verge. Not yet, but the trendlines point to it as a likely possibility.
It's already becoming a racket.
I have barely technical people I knew from college pushing the latest AI certification (I can't imagine this staying relevant for longer than 6 months) and one being hired as a frontier AI engineer.
It reminds me of 2000 all over again where knowing html got you a six-figure job.
And if management decides we don't need those 6 hours of human work, will everyone still be complaining?
It takes years to adapt fully to new tools, and it takes years for the toolmakers to figure out what the tools need to do
This is all normal. It’s also well worth the time spent learning
Just 6 hours, lol!