I was at a big tech for last 10 years, quit my job last month - I feel 50x more productive outside than inside.<p>Here is my take on AI's impact on productivity:<p>First let's review what are LLMs objectively good at:
1. Writing boiler plate code
2. Translating between two different coding languages (migration)
3. Learning new things: Summarizing knowledge, explaining concepts
4. Documentation, menial tasks<p>At a big tech product company #1 #2 #3 are not as frequent as one would think - most of the time is spent in meetings and meetings about meetings. Things move slowly - it's designed to be like that. Majority devs are working on integrating systems - whatever their manager sold to their manager and so on. The only time AI really helped me at my job was when I did a one-week hackathon. Outside of that, integrations of AI felt like more work rather than less - without much productivity boost.<p>Outside, it has proven to be a real productivity boost for me. It checks all the four boxes. Plus, I don't have to worry about legal, integrations, production bugs (eventually those will come).<p>So, depends who you are asking -- it is a huge game changer (or not).
I think a lot of people will not like to hear this but we use AI almost for everything internally. The noob way to go about this is just give it a couple of tasks and just give it complete root access to your life. That's always going to end up in disappointment. Instead, I realised, AI always needs an architect. Opinionated. Strategic. Authoritative.<p>It is quite good at following <i>most</i> orders. Hence why you must ALWAYS be in the loop. AI can augment, but not replace. Maybe some day it might. But it's not now, even with the latest SOTA models.<p>I let AI write my emails for me. But never the ability to hit send. I let AI access to my data to make informed decisions, but never let it make the final decision.<p>You may think I'm being paranoid, but I'm a very cautious person. I don't jump into new technology fresh out of the oven and this has served me well for the last 15 years. (I learned my lesson courtesy of MongoDb).<p>With AI, I am taking the same approach. Experiment, understand the limits and only then implement. Working really well so far and have managed to automate tons of tedious tasks from emails to sales to even meetings.<p>I don't use Clawdbot, not any library. I wrote my own wrappers for everything using Elixir. I used Instructor and Ash framework with Phoenix and a bunch of generators to automate tedious tasks. I control the endpoints the models are loaded from (Open router) and use a multi-model flow so no one company has enough data about me. Only bits and pieces of random user IDs.<p>Privacy is the real challenge with AI.
having been an early employee and founder of a few startups and then working at a few larger companies, most people who only ever worked at FAANG have no idea how much more productive tiny teams with ownership are.
Been a startup founder - work at Meta currently.<p>AI is making everyone faster that I’ve seen. I’d say 30% of the tickets I’ve seen in the last month have been solved by just clicking the delegate to AI button
How did you decide to work at Meta?<p>I'll be honest, just the idea of working there makes me feel like vomiting. For me, they are bizarrely evil. They're not evil like, "we're going to destroy our competition through anti competitive practices," (which they do), but "let's destroy a whole generation of minds."<p>And now with the glasses. I mean, jeeze. Can there be a stronger signal of not caring for others?<p>It's as if Meta sees people as cattle.
Though I think a lot of techies see humans as cattle, truthfully.<p>What was your rationale?<p>I guess this question is out-of-the-blue, and I don't mean for you to justify your existence, but I've never understood why people choose to work for Meta.
I feel the same - would I like a meta paycheck, sure, but I couldn't look at myself in the mirror knowing what the company I'm giving my work to does to people's brains (not just the young, though that is the most reprehensible).
I told my son I would disown him if he worked for Facebook, for the reasons stated above.<p>Then he took a contracting gig for Meta. His rationalization was that the project was an ill-specified prototype that would never see the light of day - if they wanted to throw money at him for stuff like that, he would accept it.<p>That gig is finished, and he's now thoroughly disillusioned with working for big tech.
Guess who is running product and other related functions at OpenAI and Anthropic now
From this angle, what's the difference between Meta and a junk food company?<p>Both sell things that are bad for you, but that the consumer has complete control over whether or not to consume.<p>And not all of what Meta is selling is bad. There's a lot of information exchanged on Facebook, Instagram, etc. that are good for society. Like health/nutrition advice, etc.
I've always attributed it to people being very good at convincing themselves they aren't one of the bad guys. A big paycheck makes it even easier to ignore to what you are a part of.<p>Where livelihood is concerned, rational individuals with strong morals can do irrational, and immoral things (e.g., work at the Palantir's of the world).<p>TLDR: incentives don't just shape perception, they form it
I have a theory that when you have 2 developers working in synergy, you're at something like 1.8x what 1 person can do. As you add more people you approach 2x until some point after which you start to decrease productive. I don't think that point is far beyond 5.
This is very close to the thesis, or at least theme, of the essays in The Mythical Man-Month, Fred Brooks. Some elements are dated (1975), but many feel timeless.<p>Brooks law “Adding manpower to a late software project makes it later” is just the surface of some of the metaphorical language that has most stuck with me: large systems and teams quickening entanglement in tar pits through their struggle against coordination scaling pains, conceptual integrity in design akin to preserving architectural unity of Reims cathedral, roles and limitations attempting to expand surgical teams, etc.<p>Love a good metaphor, even when its foundation is overextended or out of date. Highly recommend.
My experience of pair programming is the opposite. In a pair I get maybe 4x as much done as when working alone.<p>Mostly it's because when we hit a point where one person would get stuck, the other usually knows what to do, and we sail through almost anything with little friction.
Maybe the multiplier is 4x and by the time you have a team of ten you're back down to 2x? My theory is a bit of a hyperbole and I don't know what the multipliers would be? But I know that many times you can move quick when you're small.<p>And to your point, a single person can easily get stuck, I know that applies to me many times.
There's that but youre missing a lot of variables.
E.g. if one of you had perfect sleep and the other didn't the individual with perfect sleep will perform better for longer.<p>I don't get why people try to simplify - you're removing important details that determine performance and therefore output. This leads to false conclusions.
This. Hell even a company that is 100 people or more. Ive seen companies just grind to snails pace around 80-90 people and then still scale to 400-500 and then it's impossible to really do anything meaningful. I have tried to test for this in interviews over the years but ultimately I just end up disappointed. At this point I don't even look, just work in small independently organized groups or coops.
I love tiny teams. I hate big corp.<p>Big corporations are full with people who love to entertain 20+ people in video calls. 1-2 people speak, the other nod their heads while browsing Amazon.<p>I wouldn’t be sad if those jobs vanished.
Well, you should be terrified of those jobs vanishing I think.<p>All of these people will consequently be on the job market competing for your opportunities.<p>Yes you may feel superior to their capabilities - and may even be justified in your opinion (I know nothing about you beyond this comment)... But it'll still significantly impact your professional future if this actually happens. It would massively impact wages at the very least<p>Your viewpoint is incredibly short-sighted and not actually realizing the broad effect on the industry as a whole such a change would bring.
I agree with your categories. The majority of the usage for me is (1) and (3).<p>(1) LLMs are basically Stack Overflow on steroids. No need to go look up examples or read the documentation in most cases, spit out a mostly working starting point.<p>(3) Learning. Ramping up on an unfamiliar project by asking Antigravity questions is really useful.<p>I do think it makes devs faster, in that it takes less time to do these two things. But you're running into the 80% of the job that does not involve writing code, especially at a larger company.<p>In theory, this should allow a company to do more with fewer devs, but in reality it just means that these two activities become easier, and the 80% is still the bottleneck.
> LLMs are basically Stack Overflow on steroids<p>That, and I've never had to beg an LLM for an answer, or waste 5 minutes of my life typing up a paragraph to pre-empt the XY Problem Problem. Also never had it close my question as a duplicate of an unrelated question.<p>The accuracy tends to be somewhat lower than SO, but IMO this is a fair tradeoff to avoid having to potentially fight for an answer.
Yes, throughput is determined by the bottleneck and above a certain organization size, the bottleneck often is coordination costs.
Interesting.<p>Are you generating revenue or, otherwise, what productivity are you measuring?<p>Without generating revenue (which to be clear is a very good proxy to measure impact) everyone can be indeed very prolific in their hobbies. But labor market is about making money for a living and unless you can directly impact your day-to-day needs from your work, it can't be called productive.
Can mostly second this.<p>Working on a side project, and it's truly incredible how good AI has been for MOST of it.<p>Also, bewildering how truly awful it was at some seemingly random things - like writing not terribly difficult Assembly that mostly exists already to do Go-style hot splitting (to even get it to understand what older versions of Go did).<p>I suspect it'll still be 3 years before AI is as good at the FAANGs as it is outside, just due to the ungodly huge context and the amount of proprietary stuff it would need to learn to use effectively, plus getting all the access to it, etc.<p>But, even when it does all that, that's maybe 33% of the job.<p>I just don't see mass layoffs at the really big tech companies, unless it's more focused on just cutting and cutting than actually because people have been made redundant.<p>Even at the management level, I'm not sure we're going to see managers managing teams of 30 instead of teams of 10.<p>At the end of the day, a manager needs to know what you're doing and if you're any good at it, and there's only so many people a person can do that effectively with.<p>Maybe low-level managers go away, and it's just TLMs, but someone still needs to do your 1-on-1s and babysit those that need babysat.
2. Translating between two different coding languages (migration)<p>I have a game written in XNA<p>100% of the code is there, including all the physics that I hand-wrote.<p>All the assets are there.<p>I tried to get Gemini and Claude to do it numerous times, always with utter failure of epic proportions with anything that's actually detailed.
1 - my transition from the lobby screen into gameplay? 0% replicated on all attempts
2 - the actual physics in gameplay? 0% replicated none of it works
3 - the lobby screen itself? non-functional<p>Okay so what did it even do? Well it put together sort of a boilerplate main menu and barebones options with weird looking text that isn't what I provided (given that I provided a font file), a lobby that I had to manually adjust numerous times before it could get into gameplay, and then nonfunctional gameplay that only handles directional movement and nothing else with sort of half-working fish traveling behavior.<p>I've tried this a dozen times since 2023 with AI and as late as late last year.<p>ALL of the source code is there every single thing that could be translated to be a functional game in another language is there. It NEVER once works or even comes remotely close.<p>The entire codebase is about 20,000 lines, with maybe 3,000 of it being really important stuff.<p>So yeah I don't really think AI is "really good" at anything complex. I haven't really been proven wrong in my 4 years of using it now.
I crave to see people saying "Here's the repo btw: ..." and others trying out porting it over, just so we see all of the ways how AI fails (and how each model does) and maybe in the middle there a few ways to improve its odds. Until it eventually gets included in training data, a bit like how LLMs are oddly good at making SVGs of pelicans on bicycles nowadays.<p>And then, maybe someone slightly crazy comes along and tries seeing how much they can do with regular codegen approaches, without any LLMs in the mix, but also not manual porting.
Agreed -- coding agents / LLMs are definitely imperfect, but it's always hard to contextualize "it failed at X" without knowing exactly what X was (or how the agent was instructed to perform X)
I'm sure someone who regularly programs games in the destination language I want who also has worked with XNA in the past as a game developer could port it in a week or something yeah
- Split it in different modules / tasks<p>- Do not say: "just convert this"<p>- On critical sections you do a method-per-method-translation<p>- Dont forget: your 20.000 lines source at a whole will make any model to be distracted on longer tasks (and sessions, for sure)<p>- Do dedicated projects within Claude per each sub-module
This matches my experience. Unless it's been done to death online (crud etc) it falls on its face every time.
In those situations you basically need to guide llm to do it properly. It rarely one shots complex problems like this, especially in non web dev, but could make it faster than doing it manually.
This is a bot comment.
People who are saying they're not seeing productivity boost, can you please share where is it failing?<p>Because, I am terrified by the output I am getting while working on huge legacy codebases, it works. I described one of my workflow changes here: <a href="https://news.ycombinator.com/item?id=47271168">https://news.ycombinator.com/item?id=47271168</a> but in general compared to old way of working I am saving half of the steps consistently, whether its researching the codebase, or integrating new things, or even making fixes. I have stopped writing code, occasionally I jump into the changes proposed by LLM and make manual edits if it is feasible, otherwise I revert changes and ask it to generate again but based on my learnings from the past rejected output<p>I am terrified about what's coming
The companies laying off people have no vision. My company is a successful not for profit and we are hiring like crazy. It’s not a software company, but we have always effectively unlimited work. Why would anyone downsize because work is getting done faster? Just do more work, get more done, get better than the competition, get better at delivering your vision. We put profits back in the community and actually make life better for people. What a crazy fucking concept right?
I suspect it depends partly on how locked each individual is into a particular type of work, both skill-wise and temperamentally.<p>To give an example from a field where LLMs started causing employment worries earlier than software development: translation. Some translators made their living doing the equivalent of routine, repetitive coding tasks: translating patents, manuals, text strings for localized software, etc. Some of that work was already threatened by pre-LLM machine translation, despite its poor quality; context-aware LLMs have pretty much taken over the rest. Translators who were specialized in that type of work and too old or inflexible to move into other areas were hurt badly.<p>The potential demand for translation between languages has always been immense, and until the past few years only a tiny portion of that demand was being met. Now that translation is practically free, much more of that demand is being met, though not always well. Few people using an app or browser extension to translate between languages have much sense of what makes a good translation or of how translation can go bad. Professional translators who are able to apply their higher-level knowledge and language skills to facilitate intercultural communication in various ways can still make good money. But it requires a mindset change that can be difficult.
I'm not in translation, but a number of close friends are in the industry. Two trends I've noticed in the industry, which I think we're seeing mirrored in tech:<p>1. No one cares about quality. Even in fields you'd expect to require the 'human touch' (e.g. novel translation), publishers are replacing translators with AI. It doesn't matter if you have higher-level knowledge or skills if the company gains more from cutting your contract than it loses in sales.<p>2. Translation jobs have been replaced with jobs proofreading machine translations, which pays peanuts (since AI is 'doing most of the work') but in fact takes almost as much effort as translating from scratch (since AI is often wrong in very subtle ways). The comparison to PR reviews makes itself.
It is not entirely true that no one cares about quality. I'd like to stay optimistic and believe that those who are demanding on the quality of their production will acquire sufficient market differentiation to prevail.<p>After all, this has been Apple strategy since the 80's, and, even though there were some up's and down's, overall it's a success.
> It is not entirely true that no one cares about quality. I'd like to stay optimistic and believe that those who are demanding on the quality of their production will acquire sufficient market differentiation to prevail.<p>Maybe, but it probably requires a very strong and opinionated leader to pull off. The conventional wisdom in American business leadership seems to be to pursue the lowest level of quality you can get away with, and focus on cutting costs. And you'll have to fight that every second.<p>I don't think that's true at the individual-contributor level (pursing quality is very motivating), but they people who move up are the ones who <i>sound</i> "smart" by aping conventional wisdom.<p>> After all, this has been Apple strategy since the 80's, and, even though there were some up's and down's, overall it's a success.<p>I might give you that "since the late 90s," but there have been significant periods where that wasn't true (e.g. the early mid-90s Mac OS was buggy and had poor foundations).
someone still will, but quality will become really expensive
This is exactly right IMO. I have never worked for a company where the bottleneck was "we've run out of things to do". That said, plenty of companies run out of actual software engineering work when their product isn't competitive. But it usually isn't competitive because they haven't been able to move fast enough
That was my insight also. As a manager, you already have the headcount approved, and your people just allegedly got some significant percentage more productive. The first thought shouldn't be, great let's cut costs, it should be great now we finally have the bandwidth to deliver faster.<p>On a macro level, if you were in a rising economic tide, you would still be hiring, and turning those productivity gains into more business.<p>I wonder what the parallels are to past automations. When part producing companies moved from manual mills to CNC mills, did they fire a bunch of people or did they make more parts?
CNC machines drove down operator wages. Its similar to the translator example where the machine code is written by someone else, but the person running the machine still needs to understand. Simple pushing the go button is dangerous, being able to adapt is critical.<p>Jobs where a machinist is in charge of large chunks of the process are rarer. Large shop will have one person setting up many machines to maximize throughput.
I'm an EM as well and I've been telling my teams for a while now that I think they really only need to start worrying once our backlog starts going down instead of up. Generally, I still agree with that (and your) sentiment when you look at the long term, but in the short term, I think all of the following arguments can be made in favor of layoffs:<p>- AI tools are expensive so until the increased productivity translates to increased revenue we need to make room in the budget<p>- We expect the bottlenecks in our org to move from writing code to something else (PM or design or something) so we're cutting SWEs in anticipation of needing to move that budget elsewhere.<p>- We anticipate the skillsets needed by developers in the AI world to be fundamentally different from what they are now that it's cheaper to just lay people off, run as lean as possible, and rehire people with the skills we want in a year or two than it is to try and retrain.<p>I don't necessarily agree with those arguments (especially the last one), but I think they're somewhat valid arguments
I see similar arguments and I don't agree as well, here is why:<p>> rehire people with the skills we want in a year or two than it is to try and retrain.<p>before that future comes your company might become obsolete already, because you have lost your market share to new entrants<p>> We expect the bottlenecks in our org to move from writing code to something else<p>I would love to tell them, hey lets leverage current momentum and build, when those times come, we offer existing people with accumulated knowledge to retrain to a new type of work, if they think they're not good fit, they can leave, if they're willing, give them a chance, invest in people, make them feel safe and earn trust and loyalty from them<p>> AI tools are expensive so until the increased productivity translates to increased revenue we need to make room in the budget<p>1. Its not that expensive: 150$/seat/month -> 5 lunches? or maybe squeeze it from Sales personnel traveling with Business class?<p>2. By the time increased productivity is realized by others, company who resisted could be so far behind, that they won't be able to afford hiring engineers with those skillsets, if they think 150$ is expensive now, I am sure they will say "What??? 350k$ for this engineer?, no way, I will instead hire contractors"
business success does not scale at the speed of increased profits from layoffs.
most businesses dont actually have an infinite amount of work that has extremely high ROI. every new project at google for example has to justify the engineering spend of developing a product that has comparable margin to the ad business. Why spend 10 million a year of engineering resources on a new product that might 1. completely fail or 2. be a decent product with 20% margins when they could do nothing and keep raking in 90% margins from the ads business.
You need certain company culture, to be able to scale up, and to capture this value. Most companies can not just add new developers.<p>AI needs documentation, automation, integration tests... It works very well for remote first company, but not for in-face informal grinding approach.<p>Just year ago, client told me to delete integration tests, because "they ran too long"!
<i>>Just year ago, client told me to delete integration tests, because "they ran too long"!</i><p>Why are you surprised customers don't like spending money on the items that don't add business value. Add to that QA, documentation, security audits, etc.<p>They want to ship stuff that brings in customers and revenue day one, everything else is a cost.
> integration tests, QA etc ... the items that don't add business value<p>They absolutely do add value / prevent loss, but you need some understanding in order to see that. Not seeing it is a marker of not understanding.
<i>>They absolutely do add value </i><p>Not to the non-technical bean counters. When they allocate money they want to see you prove how that extra money translates to an immediate ROI, and it's difficult to prove that in an Excel sheet exactly what the ROI will be without making stuff up on vibes and feels.<p>Like at one German company i was at ~15 years ago, all the devs wanted a second 19" monitor on our workstations for increased productivity, and the bean counters wouldn't approve that because they wanted proof of how that expense across hundreds of people will increase our productivity and by how much %, to see if that would offset the cost.<p>This is how these people think. If you don't bring hard numbers on how much their "line will go up", they won't give you money.<p>I know this is difficult to understand from the PoV of SV Americans where gazillions of dollars just fall from the sky at their tech companies.
Does that extra work bring in more revenue? I think that’s the key question.
Because hiring less while getting more done increases margins. Your company is not for profit so doesnt care about margins. Others do.
These are words without weights. At some point the put money into software option will max out. Perhaps what we should all be doing is hiring more lawyers, there's always more legal work to be done. When you don't have weights then you can reason like this.
I’ve been screaming this too <a href="https://news.ycombinator.com/item?id=47212237">https://news.ycombinator.com/item?id=47212237</a><p>It’s refreshing to see the same sentiment from so many other people independently here.
The problem becomes if you are a service like Youtube, where you already have capture almost the entire customer base.
Yes, it's the lump of labor fallacy.<p>Doesn't exclude the possibility of short term distribution, though.
You would need to expand your capacity to find and define the work. I imagine that would be a major challenge.
> Just do more work, get more done<p>That's one of the reasons why I am terrified, because it can lead to burn out, and I personally don't like to babysit bunch of agents, because the output doesn't feel "mine", when its not "mine" I don't feel ownership.<p>And I am deliberately hitting the brake from time to time not to increase expectations, because I feel like driving someone else's car while not understanding fully how they tuned their car (even though I did those tunings by prompting)
I'm currently a product manager (was a software engineer and technical architect before), so i already lost the feeling of ownership of code. But just like when you're doing product management with a team of software engineers, testers, and UXers, with AI you can still feel ownership of the feature or capability you're shipping. So from my perspective, nothing changes regarding ownership.
It feels very much like leading a team of junior engineers or even interns who are very fast but have no idea about why we're doing anything. You have to understand the problems you're trying to solve and describe the solutions in a way they can be implemented.<p>It's not going to be written exactly like you would do it, but that's ok - because you care about the results of the solution and not its precise implementation. At some point you have to make an engineering decision whether to write it yourself for critical bits or allow the agent/junior to get a good enough result.<p>You're reviewing the code and hand editing anyway, right? You understand the specs even if your agent/junior doesn't, so you can take credit even if you didn't physically write the code. It's the same thing.
> It feels very much like leading a team of junior engineers or even interns who are very fast but have no idea about why we're doing anything<p>Yes, yes!<p>And this is problem for me, because of the pace, my brain muscles are not developing enough compared to when I was doing those things myself.<p>before, I was changing my mind while implementing the code, because I see more things while typing, and digging deeper, but now, because juniors are doing things they don't offer me a refactoring or improvements while typing the code quickly, because they obey my command instead of having "aha" moment to suggest better ways
There’s some hope that the industry will realize that managing clueless LLMs at high pace isn’t sustainable and leads to worse results, and some middle ground has to be found. Or we will reach AGI, so AI won’t be clueless anymore and really take your engineering job.
I think a lot of companies have ineffective ways to measure productivity, poor management (e.g., people who were IC's then promoted to management but have no management training or experience), incentives aren't necessarily aligned between orgs and staff, so people end up with a perverse "more headcount" means I'm better than Sandy over there. Leadership and vision have been rare in my professional life (though the corporate-owned media celebrates mediocrity in leadership all the time with puff pieces).<p>Once you get to a certain size company, this means a lot of bloat. Heck, I've seen small(ish) companies that had as many managers and administrators as ICs.<p>But You're not wrong, I'm just pointing out how an org that has 4k people can lay off a few hundred with modest impact of the financials (though extensive impact on morale).
I find LLMs are good at essentially boilerplate code. It's clear what to do and it needs to be typed in. Or areas where I really have no idea where to start, because I'm not familiar with the codebase.<p>I find anything else, I spend more time coaxing them into doing 85% of what I need that I'm better off doing it myself.<p>So they're not useless but there's only so many times in a week that I need a function to pretty-print a table in some fashion. And the code they write on anything more complex than a snippet is usually written poorly enough that it's a write-once-never-touch-again situation. If the code needs to be solid, maintainable, testable, correct (and these are kind of minimal requirements in my book) then LLMs make little impact on my productivity.<p>They're still an improvement on Google and Stack exchange, but again - only gets you so far.<p>YMMV
> I find anything else, I spend more time coaxing them into doing 85% of what I need that I'm better off doing it myself.<p>You must be working in a very niche field with very niche functionality if that's the case? I work at a company just outside of FAANG and I work in compliance. Not a terribly complex domain but very complicated scale and data integrity requirements.<p>I haven't written a single line of code manually in 2 weeks. Opus 4.6 just... works. Even if I don't give it all the context it just seems to figure things out. Occasionally it'll make an architectural error because it doesn't quite understand how the microservices interact. But these are non-trivial errors (i.e. humans could have made them as well) and when we identify such an error, we update the team-shared CLAUDE.md to make sure future agents don't repeat the error.
I often wonder what I am missing. Recently I wanted to wrap a low level vendor API with a callback API (make a request struct and request id, submit, provide a callback fn, which gets called with request IDs and messages received from vendor) to async Python (await make_request(...)). Kinda straightforward - lots of careful code of registering and unregistering callbacks, some careful thread synchronisation (callbacks get called in another thread), thinking about sane exception handling in async code. Fiddly but not rocket science.<p>What I got sort of works, as in tests pass - this with Opus 4.5. It is usable, though it doesn't exist cleanly on errors despite working to death with Claude about this. On exception it exits dirtily and crashes, which is good enough for now. I had some fancy ideas about logging messages from the vendor to be able to replay them, to be able to then reproduce errors. Opus made a real hash of it, lots of "fuck it comment out the assert so the test passes". This part is unusable and worse, pollutes the working part of the project. It made a valiant effort at mocking the vendor API for testing but really badly, instead of writing 30 lines of general code, it wrote 200 lines of inconsistent special cases that don't even work altogether. Asked to fix it it just shuffles around the special cases and gets stuck.<p>It's written messily enough that I wouldn't touch this even to remove the dead code paths. I could block a few days for it to fix but frankly in that time I can redo it all and better. So while it works I'm not gonna touch it.<p>I did everything LLM proponents say. I discussed requirements. Agent had access to the API docs and vendor samples. I said <i>think hard</i> many times. Based on this we wrote a detailed spec, then detailed inplementation plan. I hand checked a lot of the high level definitions. And yet here I am. By the time Opus went away and started coding, we had the user facing API hammered out, key implementation details (callback -> queue -> async task in source thread routing messages etc), constraints (clean handling of exceptions, threadsafe etc). Tests it has to write. Any minor detail we didn't discuss to death was coded up like a bored junior.<p>And this also wasn't my first attempt, this was attempt #3. First attempt was like, here's the docs and samples, make me a Python async API. That was a disaster. Second was more like, let's discuss, make a spec, then off you go. No good. Even just taking the last attempt time, I would have spent less time doing this by hand myself from scratch.
Just a guess, but to me it sounds like you're trying to do too much at once. When trying something like this:<p>> lots of careful code of registering and unregistering callbacks, some careful thread synchronisation (callbacks get called in another thread), thinking about sane exception handling in async code. Fiddly but not rocket science.<p>I'd expect CC to fail this when just given requirements. The way I use it is to explicitly tell it things like: "Make sure to do Y when callback X gets fired" and not "you have to be careful about thread synchronisation". "Do X, so that Exceptions are always thrown when Y happens" instead of "Make sure to implement sane Exception handling". I think you have to get a feeling for how explicit you have to get because it definitely can figure out some complexity by itself.<p>But honestly it's also requires a different way of thinking and working. It reminds me of my dad reminiscing that the skill of dictating isn't used at all anymore nowadays. Since computers, typing, or more specifically correcting what has been typed has become cheap. And the skill of being able to formulate a sentence "on the first try" is less valuable. I see some (inverse) parallel to working with AI vs writing the code yourself. When coding yourself you don't have to explicitly formulate everything you are doing. Even if you are writing code with great documentation, there's no way that it could contain all of the tacit knowledge you as the author have. At least that's how I feel working with it. I just got really started with Claude Code 2 months ago and for a greenfield project I am amazed how much I could get done. For existing, sometimes messy side projects it works a lot worse. But that's also because it's more difficult to describe explicitly what you want.
> The way I use it is to explicitly tell it things like: "Make sure to do Y when callback X gets fired" and not "you have to be careful about thread synchronisation". "Do X, so that Exceptions are always thrown when Y happens" instead of "Make sure to implement sane Exception handling".<p>At this point I'm basically programming in English, no? Trying to squeeze exact instructions into an inherently ambiguous representation. I might as well write code at this point, if this is the level of detail required. For this to work, I <i>need</i> to be able to say "make this thread-safe", maybe "by using a queue". Not explaining which synchronisation primitive to use in every last piece of the code.<p>This is my point actually. If I describe the task to accuracy level X, it still doesn't seem to work. To make it work, perhaps I need to describe it to level Y>X, but that for now takes me more time than to do it myself.<p>There's lots of variables here, how fast I am at writing code or planning structure, how close to spec the things needs to be, etc. My first "vibe code" was a personal productivity app in Claude Code, in Flutter (task timing). I have 0 idea about Dash or Flutter or any web stuff, and yet it made a complete app that did some stuff, worked on my phone, with a nice GUI, all from just a spec. From scratch, it would take me weeks.<p>...though in the end, even after 3 attempts, the final thing still didn't actually work well enough to be useful. The timer would sometimes get stuck or crash back down to 0, and froze when the app was minimised.
> At this point I'm basically programming in English, no?<p>Yea, except they can handle some degree of complexity. Its usefulness obviously really depends on that degree. And I'm sure there are still a lot of domains and types of software where that tradeoff between doing it yourself or spelling it out isn't worth it.
Based on what I've seen and heard, you have the happy path working and that's what the pro-AI people are describing with huge speedups. Figuring out and fixing the edge cases and failure modes is getting pushed into the review stage or even users, so it doesn't count towards the development time. It can even count as more speed if it generates more cases that get handled quickly.
I'm not sure I agree with this approach, or at least it doesn't work in my area. It's like self driving cars. Having 90% reliability is almost as good as 0%. I have to be confident the thing is gonna work, correctly, or at worst fail predictably.<p>I can see that there's a lot of applications where things can just randomly fail and you retry / restart, that helps with crashes.<p>But the AI can't make it <i>not crash</i>, what's to say it does the right thing when it succeeds? Again, depends on the relative cost of errors etc.
> On exception it exits dirtily and crashes, which is good enough for now<p>Silent failures and unexplained crashes are high on my list of things to avoid, but many teams just take them for granted in spite of the practical impact.<p>I think that a lot of orgs have a culture of "ship it and move on," accompanied by expectations like: QA will catch it, high turnover/lower-skill programmers commit stuff like this all the time anyway, or production code is expected to have some rough edges. I've been on teams like that, mostly in bigger orgs with high turnover and/or low engineering standards.
Two use-cases recently where Claude sucked for me:<p>1. Performance-critical code to featurize byte slices for use in a ML model. Claude kept trying to take multiple passes over the slice when the featurization can obviously be done in one. After I finally got it to do the featurization in one pass it was double-counting some bytes but not others (double counting all of them would have been fine since the feature vector gets normalized). Overall it was just very frustrating because this should have been straight-forward and instead it was dogshit.<p>2. Performance-critical code that iterates over lines of text and possibly applies transformations, similar to sed. Claude kept trying to allocate new Strings inside of the hot-loop for lines that were not transformed. When I told it to use Cow<'a, str> instead so that the untransformed lines, which make up the majority of processed lines, would not need a new allocation, Claude completely fucked up the named lifetimes. Importantly, my CLAUDE.md already tells Claude to use copy-on-write types to reduce allocations whenever possible. The agent just ignored it, which is _the_ issue with LLMs: they're non-deterministic and any guidance you provide is ultimately just a suggestion.
> I spend more time coaxing them into doing 85% of what I need that I'm better off doing it myself<p>What was the last thing you built in which you felt this was the case?
I have an app which is fairly popular. This release cycle I used Claude Code and codex to implement all the changes / features. It definitely let me move much quicker than before.<p>However now that it's in the beta stage the amount of issues and bugs is insane. I reviewed a lot of the code that went in as well. I suspect the bug fixing stage is going to take longer than the initial implementation. There are so many issues and my mental model of the codebase has severely degraded.<p>It was an interesting experiment but I don't think I would do it again this way.
The last 10% takes up 90% of the time. Usually, the time is spent finding issues you didn't even know about. This was true before LLMs.
I make mistakes when writing code, but I know what types of mistakes I make. With AI it's like a coworker who makes mistakes, sometimes they're obvious to me and sometimes they're not, because I make different mistakes.
Thanks for the insight!
Rather than trying to fix the bugs yourself, have you tried asking Claude to fix them for you?
The last time I tried AI, I tested it with a stopwatch.<p>The group used feature flags...<p><pre><code> if (a) {
// new code
} else {
// old code
}
void testOff() {
disableFlag(a);
// test it still works
}
void testOn() {
enableFlag(a);
// test it still works
}
</code></pre>
However, as with any cleanup, it doesn't happen. We have thousands of these things lying around taking up space. I thought "I can give this to the AI, it won't get bored or complain."<p>I can do one flag in ~3minutes. Code edit, pr prepped and sent.<p>The AI can do one in 10mins, but I couldn't look away. It kept trying to use find/grep to search through a huge repo to find symbols (instead of the MCP service).<p>Then it ignored instructions and didn't clean up one or the other test, left unused fields or parameters and generally made a mess.<p>Finally, I needed to review and fix the results, taking another 3-5 minutes, with no guarantee that it compiled.<p>At that point, a task that takes me 3 minutes has taken me 15.<p>Sure, it made code changes, and felt "cool", but it cost the company 5x the cost of not using the AI (before considering the token cost).<p>Even worse, the CI/CD system couldn't keep up the my individual velocity of cleaning these up, using an automated tool? Yeah, not going to be pleasant.<p>However, I need to try again, everyone's saying there was a step change in December.
I did my own experiment with Claude Code vs Cursor tab completion. The task was to convert an Excel file to a structured format. Nothing fancy at all.<p>Claude Code took 4 hours, with multiple prompts. At the end, it started to break the previous fixes in favor of new features. The code was spaghetti. There was no way I could fix it myself or steer Claude Code into fixing it the right way. Either it was a dead-end or a dice roll with every prompt.<p>Then I implemented my own version with Cursor tab completion. It took the same amount of time, 4 hours. The code had a clear object-oriented architecture, with a structure for evolution. Adding a new feature didn't require any prompts at all.<p>As a result, Claude Code was worse in terms of productivity: the same amount of time, worse quality output, no possibility of (or at best very high cost of) code evolution.
Are you able to share your prompts to Claude Code? I assume not, they are probably not saved - but this genuinely surprised me, it seems like exactly the type of task an LLM would excel at (no pun intended!). What model were you using OOI?
> this genuinely surprised me<p>Me too. After listening to all the claims about Claude Code's productivity benefits, I was surprised to get the result I got.<p>I'm not able to share details of my work. I was using Claude Opus 4.5, if I recall correctly.
The exact same prompt ? Everything depends on the prompt and it’s different tools. These days the quality and what’s build around the prompt matters as much as the code. We can’t feed generic query.
Similar happened to me just now. Claude whatever-is-the-latest-and-greatest, in Claude Code. I also tried out Windsurf's Arena Mode, with the same failure. To intercept the inevitable "holding it wrong" comments, we have all the AGENTS.md and RULES.md files and all the other snake oil you're told to include in the project. It has full context of the code, and even the ticket. It has very clear instructions on what to do (the kind of instructions I would trust an unpaid intern with, yet alone a tool marketed as the next coming of Cyber Jesus that we're paying for), in a chat with minimal context used up already. I manually review every command it runs, because I don't trust it running shell scripts unsupervised.<p>I wanted it to finish up some tests that I had already prefilled, basically all the AI had to do was convert my comments into the final assertions. A few minutes later of looping, I see it finishes and all tests are green.<p>A third of the tests were still unfilled, I guess left as an exercise for the reader. Another third was modified beyond what I told it to do, including hardcoding some things which made the test quite literally useless and the last third was fine, but because of all the miscellaneous changes it made I had to double check those <i>anyways</i>. This is about the bare minimum where I would expect these things to do good work, a simple take comment -> spit out the `assert()` block.<p>I ended up wasting more time arguing with it than if I had just done the menial task of filling out the tests myself. It sure did generate a shit ton of code though, and ran in an impressive looking loop for 5-10 minutes! And sure, the majority of the test cases were either not implemented or hardcoded so that they wouldn't actually catch a breakage, but it was all green!!<p>That's ultimately where this hype is leading us. It's a genuinely useful tool in some circumstances, but we've collectively lost the plot because untold billions have poured into these systems and we now have clueless managers and executives seeing "tests green -> code good" and making decisions based on that.
What model, what harness and about how long was your prompt to fire off this piece of work? All three matters a lot, but importantly missing from your experience.
Because its failure rate is too high. Beyond boilerplate code and CRUD apps, if I let AI run freely on the projects I maintain, I spend more time fixing its changes than if I just did it myself. It hallucinates functionally, it designs itself into corners, it does not follow my instructions, it writes too much code for simple features.<p>It’s fine at replacing what stack overflow did nearly a decade ago, but that isn’t really an improvement from my baseline.
It’s not that it just makes mistakes but it also implements things in ways I don’t like or are not relevant to the business requirements or scope of the feature / project.<p>I end up replacing any saved time with QA and code review and I really don’t see how that’s going to change.<p>In my mind I see Claude as a better search engine that understands code well enough to find answers and gain understanding faster. That’s about it.
That's my experience too. It's okay at a few things that save me some typing, but it isn't really going to do the hard work for me. I also still need to spend significant amounts of time figuring out what it did wrong and correcting it. And that's <i>frustrating</i>. I don't make those mistakes, and I really dislike being led down bad paths. If "code smells" are bad, then "AI" is a rotting corpse.
> If "code smells" are bad, then "AI" is a rotting corpse.<p>This is what's so frustrating about the hype bros for me. In most cases, everything AI spits out are code smells.<p>We're all just supposed to toss out every engineering principle we've learned all so the owner class can hire less developers and suppress wages?<p>I'm sure it's working great for everyone working on SaaS CRUD or web apps, but it's still not anywhere close to solving problems outside that sphere. Native? It's very hit and miss. It has very little design sense (because, why would it? It's a language model) so it chokes on SwiftUI, it also can't stop using deprecated stuff.<p>And that's not even that specialized. It still hallucinates cmdlets if you try to do anything with PowerShell, and has near zero knowledge about the industry I work in, a historically not tech-forward industry where things are still shared in handcrafted PDF reports emailed out to subscribers.<p>I'm going to leave this field entirely if the answer just becomes "just make everything in React/React Native because it's what the AI does best."
AI dramatically increases velocity. But is velocity productivity? Productivity relative to which scope: you, the team, the department, the company?<p>The question is really, velocity _of what_?<p>I got this from a HN comment. It really hit for me because the default mentality for engineers is to build. The more you build the better. That's not "wrong" but in a business setting it is very much necessary but not sufficient. And so whenever we think about productivity, impact, velocity, whatever measure of output, the real question is _of what_? More code? More product surface area? That was never really the problem. In fact it makes life worse majority of the time.
The real question is, is it increasing their velocity?<p>They've already admitted they just 'throw the code away and start again'.<p>I think we've got another victim of perceived productivity gains vs actual productivity drop.<p>People sitting around watching Claude churn out poor code at a slower rate than if they just wrote it themselves.<p>Don't get me wrong, great for getting you started or writing a little prototype.<p>But the code is bad, riddled with subtle bugs and if you're not rewriting it and shoving large amounts of AI code into your codebase, good luck in 6-12 months time.
Something I've been thinking about lately is if there is value in understanding the systems we produce and if we expected to?<p>If I can just vibe and shrug when someone asks why production is down globally then I'm sure the amount of features I can push out increases, but if I am still expected to understand and fix the systems I generate, I'm not convinced it's actually faster to vibe and then try to understand what's going on rather than thinking and writing.<p>In my experience the more I delegate to AI, the less I understand the results. The "slowness and thinking" might just be a feature not a bug, at times I feel that AI was simply the final straw that finally gave the nudge to lower standards.
<i>>if I can just vibe and shrug when someone asks why production is down globally</i><p>You're pretty high up in the development, decision and value-addition chain, if YOU are the responsible go-to person for these questions. AI has no impact on your position.
Naa, I'm just a programmer. Experience may vary depending on company and country, for me this has been true from tiny startups to global corporations.<p>Tangential, I don't even know what "responsible" in the corporate world means anymore, it seems to me no one is really responsible for anything. But the one thing that's almost certain is that I will fix the damn thing if I made it go boom.
It's failing when there is no data in the training set, and there are no patterns to replicate in the existing code base.<p>I can give you many, <i>many</i> examples of where it failed for me:<p>1. Efficient implementation of Union-Find: complete garbage result
2. Spark pipelines: mostly garbage
3. Fuzzer for testing something: half success, non-replicateable ("creative") part was garbage.
4. Confidential Computing (niche): complete garbage if starting from scratch, good at extracting existing abstractions and replicating existing code.<p>Where it succeeds:
1. SQL queries
2. Following more precise descriptions of what to do
3. Replicating existing code patterns<p>The pattern is very clear. Novel things, things that require deeper domain knowledge, coming up with the to-be-replicated patterns themselves, problems with little data don't work. Everything else works.<p>I believe the reason why there is a big split in the reception is because senior engineers work on problems that don't have existing solutions - LLMs are terrible at those. What they are missing is that the software and the methodology must be modified <i>in order to make the LLM work</i>. There are methodical ways to do this, but this shift in the industry is still in baby shoes, and we don't yet have a shared understanding of what this methodology is.<p>Personally I have very strong opinions on how this should be done. But I'm urging everyone to start thinking about it, perhaps even going as far as quitting if this isn't something people can pursue at their current job. The carnage is coming:/
I work as an ML engineer/researcher. When I implement a change in an experiment it usually takes at least an hour to get the results. I can use this time to implement a different experiment. Doesn't matter if I do it by hand or if I let an agent do it for me, I have enough time. Code isn't the bottleneck.<p>I also heard an opinion that since writing code is cheap, people implement things that have no economic value without really thinking it through.
+1 on the economic value line. Not everything needs to be about money but if you get paid to ship code it's about money. And now we have coworkers shipping insane amounts of "features" because it's all free to ship and being an engineer, it ends there.<p>Only it doesn't, there's product positioning, UX, information architecture, onboarding and training, support, QA, change management, analytics, reporting… sigh
> but if you get paid to ship code it's about money.<p>Tip to budding software engineers: try to not work in these sort of places, as they're about "looking busy" rather than engineering software, where the latter is where real long-lasting things are built, and the former is where startup founders spend most their money.<p>The last paragraph is where the tricky and valuable parts are, and also where AI isn't super helpful today, and where you as a human can actually help out a lot if you're just 10% better than the rest of the "engineers" who only want to ship as fast as possible.
> People who are saying they're not seeing productivity boost, can you please share where is it failing?<p>At review time.<p>There are simply too many software industries that can't delegate both authorship _and_ review to non-humans because the maintenance/use of such software, especially in libraries and backwards-compat-concerning environments, cannot justify an "ends justifies the means" approach (yet).
I don’t think the objections are not necessarily in terms of lack of productivity although my personal experience is not that of massive productivity increases. The fact that you are producing code much faster is likely just to push the bottleneck somewhere else. Software value cycles are long and complicated. What if you run into an issue in 5 years the LLM fails to diagnose or fix due to complex system interactions? How often would that happen? Would it be feasible to just generate the whole thing anew matching functionality precisely? Are you making the right architecture choices from the perspective of what the preferred modus operandi of an llm is in 5 years? We don’t know. The more experienced folks tend to be conservative as they have experienced how badly things can age. Maybe this time it’ll be different?
"it works" is a very low standard when it comes to software engineering. Why are we not holding AI generated code to the same standards as we hold our peers during code reviews?<p>I have never heard anyone say "it works" as a positive thing when reviewing code..<p>Yes, there is a productivity boost but you can't tell me there is no decrease in quality
I work in commercials.<p>We can now make 1$ million dollar commercials with 100,000$ or less. So a 90% reduction in costs - if we use AI.<p>The issue is they don’t look great. AI isn’t that great at some key details.<p>But the agencies are really trying to push for it.<p>They think this is the way back to the big flashy commercials of old. Budgets are lower than ever, and shrinking.<p>Big issue here is really the misunderstanding of cause - budgets are lower, because advertising has changed in general (TV is less and less important ) and a lot of studies showed that advertising is actually not all that effective.<p>So they are grabbing onto a lifeboat. But I’m worried there’s no land.<p>I’ve planned my exit.
> I have stopped writing code, occasionally I jump into the changes proposed by LLM and make manual edits if it is feasible, otherwise I revert changes and ask it to generate again but based on my learnings from the past rejected output<p>Isn't it a very inefficient way to learn things? Like, normally, you would learn how things work and then write the code, refining your knowledge while you are writing. Now you don't learn anything in advance, and only do so reluctantly when things break? In the end there is a codebase that no one knows how it works.
> Isn't it a very inefficient way to learn things?<p>It is. But there are 2 things:<p>1. Do I want to learn that? (if I am coming back to this topic again in 5 months, knowledge accumulates, but there is a temptation to finish the thing quickly, because it is so boring to swim in huge legacy codebase)<p>2. How long it takes to grasp it and implement the solution? If I can complete it with AI in 2 days vs on my own in 2 weeks, I probably do not want to spend too much time on this thing<p>as I mentioned in other comments, this is exactly makes me worried about future of the work I will be doing, because there is no attachment to the product in my brain, no mental models being built, no muscles trained, it feels someone else's "work", because it explores the code, it writes the code. I just judge it when I get a task
I don't know where it goes, but it sounds pretty dumb for the companies involved too. Tech companies are in the business of nurturing teams knowledgeable in things so they can build something that gives them an advantage over competition. If there is no knowledge being built, there is no advantage and no tech business.
> Tech companies are in the business of nurturing teams knowledgeable in things<p>It pains the anti-capitalist fibers in my body to say this, but no they are not. At the maximum the value is in organizational knowledge and existing assets (= source code, documentation), so that people with the least knowledge possible can make changes. In software companies in general, technical excellence and knowledge is not strongly correlated with economic success as long as you clear a certain bar (that's not that high). In comparison, in hardware/engineering companies, that's a lot more correlated.<p>In the concrete example of a legacy codebase we have here, there is even less value in trying to build up knowledge in the company, as it has already been decided that the system is to be discarded anyways.
> you would learn how things work and then write the code<p>In a legacy codebase this may require learning a lot of things about how things work just to make small changes, which may be much less efficient.
I might still be naive about the industry, but if you don't know how the legacy codebase works, you might either delegate the change to someone else in the company who does, or, if there is no one left, use this opportunity to become the person who knows at least something about it.
I asked opus 4.6 how to administer an A/B test when data is sparse. My options are to look at conversion rate, look at revenue per customer, or something else. I will get about 10-20k samples, less than that will add to cart, less than that will begin checkout, and even less than that will convert. Opus says I should look at revenue per customers. I don't know the right answer, but I know it is not to look at revenue per customers -- that will have high variance due to outlier customers who put in a large order. To be fair, I do use opus frequently, and it often gives good enough answers. But you do have to be suspicious of its responses for important decisions.<p>Edit: Ha, and the report claims it's relatively good at business and finance...<p>Edit 2: After discussion in this thread, I went back to opus and asked it to link to articles about how to handle non-normally distributed data, and it actually did link to some useful articles, and an online calculator that I believe works for my data. So I'll eat some humble pie and say my initial take was at least partially wrong here. At the same time, it was important to know the correct question to ask, and honestly if it wasn't for this thread I'm not sure I would have gotten there.
A/B tests are a statistical tool, and outliers will mess with <i>any</i> statistical measure. If your data is especially prone to that you should be using something that accounts for them, and your prompt to Opus should tell it to account for that.<p>A good way to use AI is to treat it like a brilliant junior. It knows a lot about how things work in general but very little about your specific domain. If your data has a particular shape (e.g lots of orders with a few large orders as outliers) you have to tell it that to improve the results you get back.
I did tell it that I expect to see something like a power-law distribution in order value, so I think I pretty much followed your instructions here. Btw, if you do know the right thing to do in my scenario, I'd love to figure it out. This is not my area of expertise, and just figuring it out through articles so far.
I recommend reading Wikipedia and talking to LLMs to get this one. Order values do follow power-law distributions (you're probably looking for an exponential or a Zipf distribution.) You want to ask how to perform a statistical test using these distributions. I'm a fan of Bayesian techniques here, but it's up to you if you want to use a frequentist approach. If you can follow some basic calculus you can follow the math for constructing these statistical tests, if not some searching will help you find the formulas you need.
Thanks for the suggestions! I didn't want to do the math myself, but I did take your suggestion and found some articles discussing ways to make it work even with a non-normal distribution:<p>- <a href="https://cxl.com/blog/outliers/" rel="nofollow">https://cxl.com/blog/outliers/</a><p>- <a href="https://www.blastx.com/insights/the-best-revenue-significance-calculator-for-a-b-testing" rel="nofollow">https://www.blastx.com/insights/the-best-revenue-significanc...</a><p>- (online tool to calculate significance) <a href="https://www.blastx.com/rpv-calculator" rel="nofollow">https://www.blastx.com/rpv-calculator</a><p>I'm not checking their math, but the articles make sense to me, and I trust they did implement it correctly. In the end the LLM did get me to the correct answer by suggesting the articles, so I guess I should eat some humble pie and say it _did_ help me. At the same time, if I didn't have the intuition that using rpv as-is in a t-test would be noisy, and the suggestions from this comment thread, I think I could have gone down the wrong path. So I'm not sure what my conclusion is -- maybe something like LLMs are helpful once you ask the right question.
One heuristic I like to use when thinking about this question (and I honestly wish the answer space here were less emotionally charged, so we could all learn from each other) is that: LLMs need a human to understand the shape of the solution to check the LLM's work. In fields that I have confirmed expertise in, I can easily nudge and steer the LLM and only skim its output quickly to know if it's right or wrong. In fields I don't, I first ask the LLM for resources (papers, textbooks, articles, etc) and familiarize myself with some initial literature first. I then work with the LLMs slowly to make a solution. I've found that to work well so far.<p>(I also just love statistics and think it's some of the most applicable math to everyday life in everything from bus arrival times to road traffic to order values to financial markets.)
I don't want to generalize from my specific situation too much, but I want to offer an anecdote from my neck of the woods. On my personal sub, I agree it is kinda crazy the kind of projects I can get into now with little to no prior knowledge.<p>On the other hand, our corporate AI is.. not great atm. It was briefly kinda decent and then suddenly it kinda degraded. Worst case is, no one is communicating with us so we don't know what was changed. It is possible companies are already trying to 'optimize'.<p>I know it is not exactly what you are asking. You are saying capability is there, but I am personally starting to see a crack in corporate willingness to spend.
Sometimes I’m scared.<p>Sometimes I realise that this particular task has been <i>slower</i> than if I’d done it myself when I take in to account full wall clock time.<p>I can’t tell what type of task is going to work ahead of time yet.
Basically it tricks you into making the code less maintainable, so while it <i>seems</i> to boost productivity, it's just delaying huge failures.
<p><pre><code> People who are saying they're not seeing productivity boost, can you please share where is it failing?
</code></pre>
Believe it or not, I still know many devs who do not use any agents. They're still using free ChatGPT copy and paste.<p>I'm going to guess that many people on HN are also on the "free ChatGPT isn't that good at programming" train.
Which one would you recommend as the best right now? Claude Code?
Not everyone has the capability to rent out data center tier hardware to just do their job. These things require so much damn compute you need some serious heft to actually daisy chain enough stages either in parallel or deep to get enough tokens/sec for the experience to go ham. If you're making bags o' coke money, and deciding to fund Altman's, Zuckernut's or Amazon/Google's/Microsoft's datacenter build out, that's on you. Rest of us are just trying to get by on bits and bobs we've kept limping over the years. If opencode is anything to judge the vibecoded scene by, I'm fairly sure at some point the vibe crowd will learn the lesson of isolating the <i>most expensive computation ever</i> from the hot loop, then maybe find one day all they needed was maybe something to let the model build a context, and a text editor.<p>Til then wtf_are_these_abstractions.jpg
This is my current problem: I can get work to pay for a Claude Max subscription, but for personal use or to learn how to use it that's a big price tag.<p>I worry that we're returning to an era of renting core development tools. After the huge benefits from free and open source tools, that's a bitter pill to swallow.
> They're still using free ChatGPT copy and paste<p>Probably that's the reason why some people are sure their job is still safe.<p>Nature of job is changing rapidly
When it comes to novel work, LLMs become "fast typers" for me and little more. They accelerate testing phases but that's it. The bar for novelty isn't very high either - "make this specific system scale in a way that others won't" isn't a thing an LLM can <i>ever</i> do on its own, though it can be an aid.<p>LLMs also are quite bad for security. They can find simple bugs, but they don't find the really interesting ones that leverage "gap between mental model and implementation" or "combination of features and bugs" etc, which is where most of the interesting security work is imo.
What was your take on this?<p><a href="https://aisle.com/blog/what-ai-security-research-looks-like-when-it-works" rel="nofollow">https://aisle.com/blog/what-ai-security-research-looks-like-...</a>
I think your analysis is a bit outdated these days or you may be holding it wrong.<p>I am doing novel work with codex but it does need some prompting ie. exploring possibilities from current codebase, adding papers to prompt etc.<p>For security, I think I generally start a new thread before committing to review from security pov.
You can do novel work with an LLM. <i>You</i> can. The LLM can't. It can be an aid - exploring papers, gathering information, helping to validate, etc. It can't do the actual <i>novel</i> part, fundamentally it is limited to what it is trained on.<p>If you are relying on the LLM and context, then unless your context is a secret your competitor is only ever one prompt behind you. If you're willing to pursue true novelty, you need a human and you can leap beyond your competition.
This is basically my take as well!
Same. Whenever an article like this one pops up the comments seem to be filled with confirmation bias. People who don't see a productivity boost agree with the article.<p>I work at tech company just outside of big tech and I feel fairly confident that we won't have a need for the amount of developers we currently have within 3-4 years.<p>The bottleneck right now is reviewing and I think it's just a matter of time before our leadership removes the requirement for human code reviews (I am already seeing signs of this ("Maybe for code behind feature flags we don't need code reviews?").<p>Whenever there's an incident, there is a pagerduty trigger to an agent looking at the metrics, logs, software component graphs, and gives you an hypothesis on what the incident is due to. When I push a branch with test failures, I get one-click buttons in my PR to append commits fixing those tests failures (i.e. an agent analyses the code, the jira ticket, the tests, etc. and suggests a fix for the tests failing). We have a Slack agent we can ping in trivial feature requests (or bugs) in our support channels.<p>The agents are being integrated at every step. And it's not like the agents will stop improving. The difference between GPT3.5 and Opus 4.6 is so massive. So what will the models look like in 5 years from now?<p>We're cooked and the easiest way to tell someone works at a company who hasn't come very far in their AI journey is that they're not worried.
I feel like this might be heavily dependent on both your task and the AI you're using? What language do you code in and what AI do you use? And are your tasks pretty typical/boilerplate-y with prior art to go off of, or novel/at-the-edge-of-tech?
It’s been my experience as of recently. I point it at an issue tracker and ask it to investigate, write a test to reproduce the problem and plan a fix together. There’s lots of hand holding from me but it saves me a lot of work and I’ve been surprised by its comfort with legacy code bases. For now I feel empowered, and I’m actually working more intensively, but I was wondering to myself if I’m going run out of work this year. Interestingly, our metrics show that output is slowed by increased workload on reviewers.
I think you can more stuff done earlier but the quality is not good or it doesn't work as expected if you tinker with it enough.
Fixing the issues from the generated code usually doesn't work at all
I'm with you. The project I'm working on is moving at phenomenal velocity. I'm basically spending my time writing specs and performing code reviews. As long as my code review comments and design docs are clear I get a secure, scalable, and resilient system.<p>Tests were always important, but now they are the gatekeepers to velocity.
Outside of coding/non-physical areas, the impact can be quite muted. I haven't seen much impact on surgical procedures, for example (but maybe others have?).
I see an individual productivity boost, but not necessarily a collective one.<p>I don't think features per hour is really what is holding back most established businesses.<p>My experiences suggest that we still have some time before the people that understand the plumbing of the business _and_ AI bubble up to positions of authority through wielding it practically and successfully at increasingly greater scale.
I’m currently working across like 5 projects (was 4 last week but you know how it is). I now do more in days than others might in a week.<p>Yesterday a colleague didn’t quite manage to implement a loading container with a Vue directive instead of DOM hacks, it was easier for me to just throw AI at the problem and produced a working and tested solution and developer docs than to have a similarly long meeting and have them iterate for hours.<p>Then I got back to training a CNN to recognize crops from space (ploughing and mowing will need to be estimated alongside inference, since no markers in training data but can look at BSI changes for example), deployed a new version of an Ollama/OpenAI/Anthropic proxy that can work with AWS Bedrock and updated the docs site instructions, deployed a new app that will have a standup bot and on-demand AI code review (LiteLLM and Django) and am working on codegen to migrate some Oracle forms that have been stagnating otherwise.<p>It’s not funny how overworked I am and sure I still have to babysit parallel Claude Code sessions and sometimes test things manually and write out changes, but this is a completely different work compared to two or three years ago.<p>Maybe the problem spaces I’m dealing with are nothing novel, but I assume most devs are like that - and I’d be surprised at people’s productivity not increasing.<p>When people nag in meetings about needing to change something in a codebase, or not knowing how to implement something and its value add, I’ll often have something working shortly after the meeting is over (due to starting during it).<p>Instead of sending adding Vitest to the backlog graveyard, I had it integrated and running in one or two evenings with about 1200 tests (and fixed some bugs). Instead of talking about hypothetical Oxlint and Oxfmt performance improvements, I had both benchmarked against ESLint and Prettier within the hour.<p>Same for making server config changes with Ansible that I previously didn’t due to additional friction - it is mostly just gone (as long as I allow some free time planned in case things vet fucked up and I need to fix them).<p>Edit: oh and in my free time I built a Whisper + VLM + LLM pipeline based on OpenVINO so that I can feed it hours long stream VODs and get an EDL cut to desired length that I can then import in DaVinci Resolve and work on video editing after the first <i>basic</i> editing prepass is done (also PyScene detect and some audio alignment to prevent bad cuts). And then I integrated it with subscription Claude Code, not just LiteLLM and cloud providers with per-token costs for the actual cuts making part (scene description and audio transcriptions stay local since those don't need a complex LLM, but can use cloud for cuts).<p>Oh and I'm moving from my Contabo VPSes to running stuff inside of a Hetzner Server Auction server that now has Proxmox and VMs in that, except this time around I'm moving over to Ansible for managing it instead of manual scripts as well, and also I'm migrating over from Docker Swarm to regular Docker Compose + Tailscale networks (maybe Headscale later) and also using more upstream containers where needed instead of trying to build all of mine myself, since storage isn't a problem and consistency isn't that important. At the same time I also migrated from Drone CI to Woodpecker CI and from Nexus to Gitea Packages, since I'm already using Gitea and since Nexus is a maintenance burden.<p>If this becomes the new “normal” in regards to everyone’s productivity though, there will be an insane amount of burnout and devaluation of work.
> When people nag in meetings about needing to change something in a codebase, or not knowing how to implement something and its value add, I’ll often have something working shortly after the meeting is over (due to starting during it).<p>We've started building harnesses to allow people who don't understand code to create PRs to implement their little nags. We rely on an engineer to review, merge, and steward the change but it means that non-eng folks do not rely on us as a gate. (We're a startup and can't really afford "teams" to do this hand-holding and triage for us.)<p>As you say we're all a bit overworked and burned out. I've been context switching so much that on days when I'm very productive I've started just getting headaches. I'm achieving a lot more than before but holding the various threads in my head and context switching is just a <i>lot</i>.
>I now do more in days than others might in a week.<p>I've always done more in days than others might in a week. YMMV.
So do I, this is why I work 15 hours a week [1] and laugh at those that use this new productivity tool to work themselves even harder for the same pay. Wasn’t the point of automation to work less?<p>1: pre-AI. Not keen on becoming a manager of an idiot savant, so I’m planning my exit.
> I am terrified about what's coming<p>Why? This is great. AI fixing up huge legacy codebases is just taking the jobs humans would never want to do.
A terminology tangent because it's an econ publication: Notice that the article doesn't talk about productivity.<p>Productivity is a term of art in economics and means you generate more units of output (for example per person, per input, per wages paid) but doesn't take quality or otherwise desireability into account. It's best suited for commodities and industrial outputs (and maybe slop?).
Meanwhile gemini tells me my go code doesn't compile (it does)<p>Gaslight me by telling me I must be a time traveler because I use go 1.26 but the latest version actually is 1.24<p>And tell me I can't use wg.Go() because this function does not exist (it does)
> my job is easier now, I do less.
> I am terrified about what's coming.<p>God I hope I never ever have to work with you
I can only explain it by people not having used Agentic tools and or only having tried it 9 months ago for a day before giving up or having such strict coding style preferences they burn time adjusting generated code to their preferences and blaming the AI even though they’re non-functional changes and they didn’t bother to encode them into rules.<p>The productivity gains are blatantly obvious at this point. Even in large distributed code bases. From jr to senior engineer.
You were probably deficient in RESEARCH skills before. No offense to you, since I was also like this once. LLMs research and put the results on the plate. Yes, for people who were deficient in research skills, I can see 2-3x improvements.<p>Note1: I have "expert" level research skills. But LLMs still help me in research, but the boost is probably 1.2x max. But<p>Note2: By research, I mean googling, github search, forum search, etc. And quickly testing using jsfiddle/codepen, etc.
no worries, I do not get offended quickly.<p>But I also think you are overestimating your RESEARCH skills, even if you are very good at research, I am sure you can't read 25 files in parallel, summarize them (even if its missing some details) in 1 minute and then come up with somewhat working solution in the next 2 minutes.<p>I am pretty sure, humans can't comprehend reading 25 code files with each having at least 400 lines of non-boilerplate code in 2 minutes. LLM can do it and its very very good at summarizing.<p>I can even steer its summarizing skills by prompting where to focus on when its reading files (because now I can iterate 2-3 times for each RESEARCH task and improve my next attempt based on shortcomings in the previous attempt)
OK, it's not just RESEARCH, but "RESEARCHability" of the source content [in this case code], and also critical analysis ability [not saying you are deficient in anything, speaking in general terms].<p>In this example, if the 25 files are organized nicely, and I had I nice IDE that listed class/namespace members of each file neatly, I might take 30 minutes to understand the overall structure.<p>Morever, If I critically analyzed this, I would ask "how many times does this event of summarizing 25 files happen"? I mean, are we changing codebases every day? No, it's a one time cost. Moreover, manually going through will provide insight not returned by LLM.<p>Obviously, every case is different, and perhaps you do need to RESEARCH new codebases often, I dunno!
Ok Mr. Expert Level Researcher, go back and read the comment of parent again to find out that it has nothing to do with deficiency in research skills.
please don't change your comment constantly (or at maybe add UPDATE 1/2/3), because you had different words before, like you were saying something in this fashion:<p>* you probably lack good RESEARCH skills<p>* I can see at most 1.25x improvements - now it is 2-3x<p>By updating your comment you are making my reply irrelevant to your past response
I don't write code for a living but I administer and maintain it.<p>Every time I say this people get really angry, but: so far AI has had almost no impact on my job. Neither my dev team nor my vendors are getting me software faster than they were two years ago. Docker had a bigger impact on the pipeline to me than AI has.<p>Maybe this will change, but until it does I'm mostly watching bemusedly.
Yep. All AI has done for me is give me the power of how good search engines were 10+ years ago, where I could search for something and find actually relevant and helpful info quickly.<p>I've seen lots of people say AI can basically code a project for them. Maybe it can, but that seems to heavily depend on the field. Other than boilerplate code or very generic projects, it's a step above useless imo when it comes to gamedev. It's about as useful as a guy who read some documentation for an engine a couple years ago and kind of remembers it but not quite and makes lots of mistakes. The best it can do is point me in the general direction I need to go, but it'll hallucinate basic functions and mess up any sort of logic.
My experience is the same. There are modest gains compensating for lack of good documentation and the like, but the human bottlenecks in the process aren't useless bureaucracy. Whether or not a feature or a particular UX implementation of it makes sense, these things can't be skipped, sped up or handed off to any AI.
Thinking of it, I haven’t seen as many “copy paste from StackOverflow” memes lately. Maybe LLMs have given people the ability to<p>1) Do that inside their IDEs, which is less funny<p>2) Generate blog post about it instead of memes
> All AI has done for me is give me the power of how good search engines were 10+ years ago<p>So the good old days before search engines were drowning with ads and dark patterns. My assumption is big LLMs will go in the same direction after market capture is complete and they need to start turning a profit. If we are lucky the open source models can keep up.
What language/engine did you try it with for gamedev? Just curious if it was weak in a popular engine.
It makes me wonder if the majority of all-in on AI folks are quite young and never experienced pre-enshittification search.
> how good search engines were 10+ years ago<p>For me this is a huge boost in productivity. If I remember how I was working in the past (example of Google integration):<p>Before:<p><pre><code> * go through docs to understand how to start (quick start) and things to know
* start boilerplate (e.g. install the scripts/libs)
* figure out configs to enable in GCP console
* integrate basic API and test
* of course it fails, because its Google API, so difficult to work with
* along the way figure out why Python lib is failing to install, oh version mismatch, ohh gcc not installed, ohh libffmpeg is required,...
* somehow copy paste and integrate first basic API
* prepare for production, ohhh production requires different type of Auth flow
* deploy, redeploy, fix, deploy, redeploy
* 3 days later -> finally hello world is working
</code></pre>
Now:<p><pre><code> * Hey my LLM buddy, I want to integrate Google API, where do I start, come up with a plan
* Enable things which requires manual intervention
* In the meantime LLM integrates the code, install lib, asks me to approve installation of libpg, libffmpeg,....
* test, if fails, feed the error back to LLM + prompt to fix it
* deploy</code></pre>
Are you using Claude Opus 4.5/6?<p>If not, then you’re not close to the cutting edge.
Same here, more or less, in the ops world. Yeah, I use AI but I can't honestly say it's massively improved my productivity or drastically changed my job in any way other than the emails I get from the other managers at my work are now clearly written by AI.<p>I can turn out some scripts a little bit quicker, or find an answer to something a little quicker than googling, but I'm still waiting on others most of the time, the overall company processes haven't improved or gotten more efficient. The same blockers as always still exist.<p>Like you said, there has been other tech that has changed my job over time more than AI has. The move to the cloud, Docker, Terraform, Ansible, etc. have all had far more of an impact on my job. I see literally zero change in the output of others, both internally and externally.<p>So either this is a massively overblown bubble, or I'm just missing something.
You're missing something.<p>I've been in ops for 30 years, Claude Code has changed how I work. Ops-related scripting seems to be a real sweet spot for the LLMs, especially as they tend to be smaller tools working together. It can convert a few sentences into working code in 15-30 minutes while you do something else. I've given it access to my apache logs Elastic cluster, and it does a great job at analyzing them ("We suspect this user has been compromised, can you find evidence of that?"). It's quite startling, actually, what it's able to do.
Yeah, it's useful for scripting, but it's still only marginally faster. It certainly hasn't been "groundbreaking productivity" like it's being sold.<p>The problem with analyzing logs is determinism. If I ask Claude to look for evidence of compromise, I can't trust the output without also going and verifying myself. It's now an extra step, for what? I still have to go into Elastic and run the actual queries to verify what Claude said. A saved Kibana search is faster, and more importantly, deterministic. I'm not going to leave something like finding evidence of compromise up to an LLM that can, and does, hallucinate especially when you fill the context up with a ton of logs.<p>An auditor isn't going to buy "But Claude said everything was fine."<p>Is AI actually finding things your SIEM rules were missing? Because otherwise, I just don't see the value in having a natural language interface for queries I already know how to run, it's less intuitive for me and non deterministic.<p>It's certainly a useful tool, there's no arguing that. I wouldn't want to go back to working with out it. But, I don't buy that it's already this huge labor market transformation force that's magically 100x everyone's productivity. That part is 100% pure hype, not reality.
The tolerance for indeterminacy is I think a generational marker; people ~20 years younger than me just kind of think of all software as indeterminate to begin with (because it's always been ridiculously complicated and event-driven for them), and it makes talking about this difficult.
I shudder to think of how many layers of dependency we will one day sit upon. But when you think about it, aren’t biological systems kind of like this too? Fallible, indeterminable, massive, labyrinthine, and capable of immensely complex and awe inspiring things at the same time…
People younger than me are not even adults. I grew up during the dial up era and then the transition to broadband. I don't think software is indeterminate.
>still only marginally faster.<p>Is it? A couple days ago I had it build tooling for a one-off task I need to run, it wrote ~800 lines of Python to accomplish this, in <30m. I found it was too slow, so I got it to convert it to run multiple tasks in parallel in another prompt. Would have taken a couple days for me to build from hand, given the number of interruptions I have in the average day. This isn't a one-off, it's happening all the time.
<i>> ... but I'm still waiting on others most of the time, the overall company processes haven't improved or gotten more efficient. The same blockers as always still exist.</i><p>And that's the key problem, isn't it? I maintain current organizations have the "wrong shape" to fully leverage AI. Imagine instead of the scope of your current ownership, you own everything your team or your whole department owns. Consider what that would do to the meetings and dependencies and processes and tickets and blockers and other bureaucracy, something I call "Conway Overhead."<p>Now imagine that playing out across multiple roles, i.e. you also take on product and design. Imagine what that would do to your company org chart.<p>I added a much more detailed comment here: <a href="https://news.ycombinator.com/item?id=47270142">https://news.ycombinator.com/item?id=47270142</a>
> Imagine instead of<p>> Now imagine<p>> Imagine what that would do<p>Imagine if your grandma had wheels! She'd be a bicycle. Now imagine she had an engine. She could be a motorcycle! Unfortunately for grandma, she lives in reality and is not actually a motorcycle, which would be cool as hell. Our imagination can only take us so far.<p>To more substantively reply to your longer linked comment: your hypothesis is that people spend as little as 10% of time coding and the other 90% of time in meetings, but that if they could code more, they wouldn't need to meet other people because they could do all the work of an entire team themselves[1]. The problem with your hypothesis is that you take for granted that LLMs <i>actually</i> allow people to do the work of an entire team themselves, and that it is merely bureacracy holding them back. There have been absolutely zero indicators that this is true. No productivity studies of individual developers tackling tasks show a 10x speedup; results tend to be anywhere from +20% to <i>minus</i> 20%. We aren't seeing amazing software being built by individual developers using LLMs. There is still only one Fabrice Bellard in the world, even though if your premise could escape the containment zone of imagination anyone should be able to be a Bellard on their own time with the help of LLMs.<p>[1] Also, this is basically already true without LLMs. It is the reason startups are able to disrupt corporate behemoths. If you have just a small handful of people who spend the majority of their work time writing code (by hand! No LLMs required!), they can build amazing new products that outcompete products funded by trillion-dollar entities. Your observation of more coding = less meetings required in the first place has an element of truth to it, but not because LLMs are related to it in any particular way.
<p><pre><code> > Imagine if your grandma had wheels! She'd be a bicycle.
</code></pre>
I always took this to be a sharp jab saying the entire village is riding your grandma, giving it a very aggressive undertone. It's pretty funny nonetheless.<p>Too early to say what AI brings to the efficiency table I think. In some major things I do it's a 1000x speed up. In others it is more a different way of approaching a problem than a speed up. In yet others, it is a bit of an impediment. It works best when you learn to quickly recognize patterns and whether it will help. I don't know how people who are raised with ai will navigate and leverage it, which is the real long-term question (just as the difference between pre- and post-smartphone generations is a thing).
<i>> No productivity studies of individual developers tackling tasks show a 10x speedup; results tend to be anywhere from +20% to minus 20%.</i><p>The only study showing a -20% came back and said, "we now think it's +9% - +38%, but we can't prove rigorously because developers don't want to work without AI anymore": <a href="https://news.ycombinator.com/item?id=47142078">https://news.ycombinator.com/item?id=47142078</a><p>Even at the time of the original study, most other rigorous studies showed -5% (for legacy projects, obsolete languages) to 30% (more typical greenfield AND brownfield projects) way back in 2024. Today I hear numbers up to 60% from reports like DX.<p>But this is exactly missing the point. Most of them are still doing things the old way, including the very process of writing code. Which brings me to this point:<p><i>> There have been absolutely zero indicators that this is true.</i><p>I could tell you my personal experience, or link various comments on HN, or point you to blogs like <a href="https://ghuntley.com/real/" rel="nofollow">https://ghuntley.com/real/</a> (which also talks about the origanizational impedance mismatch for AI), but actual code would be a better data point.<p>So there are some open-source projects worth looking at, but they are typically dismissed because they look so weird to us. Here's two mostly vibe-coded (as in, minimal code review, apparently) projects that people shredded for having weird code, but is already used by 10s of 1000s of people, up to 11 - 18K stars now. Look at the commit volume and patterns for O(300K) LoC in a couple of months, mostly from one guy and his agent:<p><a href="https://github.com/steveyegge/beads/graphs/commit-activity" rel="nofollow">https://github.com/steveyegge/beads/graphs/commit-activity</a><p><a href="https://github.com/steveyegge/gastown/graphs/commit-activity" rel="nofollow">https://github.com/steveyegge/gastown/graphs/commit-activity</a><p>It's like nothing we've seen before, almost equal number of LoC additions and deletions, in the 100s of Ks! It's still not clear how this will pan out long term, but the volume of code and apparent utility (based purely on popularity) is undeniable.
> we now think it's +9% - +38%<p>If you are referring to the following quote [0], you are off by a sign:<p>> we now estimate a speedup of -18% with a confidence interval between -38% and +9%.<p>[0] <a href="https://metr.org/blog/2026-02-24-uplift-update/" rel="nofollow">https://metr.org/blog/2026-02-24-uplift-update/</a>
That update blog is funny. The only data they can get at reports slowdowns, but they struggle to believe it because developers self-report amazing speedups.<p>You'd get the same sort of results if you were studying the benefits of substance abuse.<p>"It is difficult to study the downsides of opiates because none of our participants were willing to go a day without opiates. For this reason, opiates must be really good and we're just missing something."
My bad, I messed up by being lazy while switching from decreases in time taken (that they report) to increased in throughput. (Yes, it's not just flipping the sign, but as I said, I was being lazy!) The broad point still holds, their initial findings have been reversed, and they expect selection effects masked a higher speedup.<p>The language is confusing, but the chart helps: <a href="https://metr.org/assets/images/uplift-2026-post/uplift_timeline.png" rel="nofollow">https://metr.org/assets/images/uplift-2026-post/uplift_timel...</a>
> they are typically dismissed because they look so weird to us.<p>I dismiss them because Yegge's work (if it can even be called his work, given that he doesn't look at the code) is steaming garbage with zero real-world utility, not "because they look weird". You suggest the apparent utility is undeniable, while saying "based purely on popularity" -- but popularity is in no way a measure of utility. Yegge is a conman who profited hundreds of thousands of dollars shilling a memecoin rugpull tied to these projects. The actual thousands of users are people joining the hypetrain, looking to get in on the promised pyramid scheme of free money where AI will build the next million dollar software for you, if only you have the right combination of .md files to make it work. None of these software are actually materialising, so all the people in this bubble can do is make more AI wrappers that promise to make other AI wrappers that will totally make them money.<p>I am completely open to being proven wrong by a vibe-coded open source application that is <i>actually useful</i>, but I haven't seen a single one. Literally not even one. I would count literally anything where the end-product is not an AI wrapper itself, which has tens to hundreds of thousands of users, and which was written entirely by agents. One example of that would be great. Just one. There have been a couple of attempts at a web browser, and Claude's C compiler, but neither are actually useful or have any real users; they are just proofs of concept and I have seen nothing that convinces me they are a solid foundation from which you could actually build useful software from, or that models will ever be on a trajectory to make them actually useful.
The memecoin thing was stupid, totally. Yegge should never have touched it, because well, crypto, but also because that's a distraction from the actual project.<p><i>> popularity is in no way a measure of utility</i><p>Why would it be popular if it's not useful? Yegge is not like some superstar whose products are popular just because he made them. And while some people may be chasing dollars, most of them are building software that scratches an itch. (Search for Beads on GitHub, you'll find thousands of public repos, and lord knows how many private repos.)<p>Beads has certainly made my agents much more effective, even the older models. To understand its utility you have to do agentic coding for a while, see the stupid mistakes agents make because they forget everything, and then introduce Beads and see almost all those issues melt away.<p><i>> None of these software are actually materialising</i><p>They are if you look for them. There are many indications (often discussed here) showing spikes in apps on app stores, number of GitHub projects, and Show HN entries. Now, you may dismiss these as "not actually useful", and at this volume that's undoubtedly true for a lot of them.<p>But there is already early data showing growth not only in mobile app downloads, but also time spent per user and <i>revenue</i> -- which are pretty clear indications of utility: <a href="https://sensortower.com/blog/state-of-mobile-2026" rel="nofollow">https://sensortower.com/blog/state-of-mobile-2026</a><p>Edit: it occurs to me that by "vibe-coding" we may be talking about two different things -- I tend to mean "heavily AI-assisted coding" whereas you likely mean "never look at the code YOLO coding." I'll totally agree that YOLO vibe-coded apps by non-experts will be crap. Other than Beads and Gastown I don't know of any such app that is non-trivial. But then those were steered by a highly experienced engineer, and my original point was, vibe-coding correctly could look very weird by today's best practices.
> I tend to mean "heavily AI-assisted coding" whereas you likely mean "never look at the code YOLO coding."<p>The original point that sparked this sub-thread though is that AI is being overhyped. If actual vibe coding (YOLO it, never look at or understand the code, thus truly enabling non-technical folk to have revolutionary power and ability) doesn't work, then AI is yet just another tool in the toolbelt like any other developer life enhancing tech we've had so far, it's just a new form of IDE.<p>Being a new form of IDE, while very useful, isn't exactly entire economy transforming revolutionary tech. If it can't be used by someone with zero computer/eng experience to build something useful and revenue generating, the amount of investment we've seen into it is way overblown and is well overdue for a pretty severe correction.<p>I buy AI as a "developer enhancing tool" just like any other devtools that we've seen over my career. I don't currently buy it as a "total labor economy transformation force."
This isn't the counter you think it is. It's too much to expect existing behemoths to reshape their orgs substantially on a quick enough timeline. The gains will be first seen in new companies and new organizations, and they will be able to stay flat a longer and outcompete the behemoths.
What a load of fluff lmao. Are you Nadella?
Ops hasn't been in the crosshairs of Ai yet.<p>Imo it's only a matter of time as companies start to figure out how to use ai. Companies don't seem to have real plans yet and everyone is figuring out ai in general out.<p>Soon though I will think agents start popping up, things like first line response to pages, executing automation
Youre not missing anything.<p>Humans are funny. But most cant seem to understand that the tool is a mirage and they are putting false expectations on it. E.g. management of firms cutting back on hiring under the expectation that LLMs will do magic - with many cheering 'this is the worst itll be bro!!".<p>I just hope more people realise before Anthropic and OAI can IPO. I would wager they are in the process of cleaning up their financials for it.
This is a classic case of Productivity Paradox when personal computers were first introduced into workplaces in the 80s.<p>A famous economist once said, "You can see the computer age everywhere but in the productivity statistics."<p>There are many reasons for the lag in productivity gain but it certainly will come.<p><a href="https://en.wikipedia.org/wiki/Productivity_paradox" rel="nofollow">https://en.wikipedia.org/wiki/Productivity_paradox</a>
That's only certain if investments in tech infrastructure <i>always</i> led to productivity increases. But sometimes they just don't. Lots of firms spent a lot of money on blockchain five years ago, for instance, and that money is just gone now.
I find it odd the universal assumption that AI is going to be good for productivity<p>The loss of skills, complete loss of visibility and experience with the codebase, and the complete lack of software architecture design, seems like a massive killer in the long term<p>I have a feeling that we're going to see productivity with AI drop through the floor
I'd claim the opposite. Better models design better software, and quickly better software than what most software developers were writing.<p>Just yesterday I asked Opus 4.6 what I could do to make an old macOS AppKit project more testable, too lazy to even encumber the question with my own preferences like I usually do, and it pitched a refactor into Elm architecture. And then it did the refactor while I took a piss.<p>The idea that AI writes bad software or can't improve existing software in substantial ways is really outdated. Just consider how most human-written software is untested despite everyone agreeing testing is a good idea simply because test-friendly arch takes a lot of thought and test maintenance slow you down. AI will do all of that, just mention something about 'testability' in AGENTS.md.
OK so this comes back to the question I started this subthread with: where is this better software? Why isn't someone selling it to me? I've been told for a year it's coming any day now (though invariably the next month I'm told last month's tools were in fact crap and useless compared to the new generation so I just have to wait for this round to kick in) and at some point I do have to actually see it if you expect me to believe it's real.
How would you know if all software written in the last six months shipped X% faster and was Y% better?<p>Why would you think you have your finger on the pulse of general software trends like that when you use the same, what, dozen apps every week?<p>Just looking at my own productivity, as mere sideprojects <i>this month</i>, I've shipped my own terminal app (replaced iTerm2), btrfs+luks NAS system manager, overhauled my macOS gamepad mapper for the app store, and more. All fully tested and really polished, yet I didn't write any code by hand. I would have done none of that this month without AI.<p>You'd need some real empirics to pick up productivity stories like mine across the software world, not vibes.
Right, I'm sympathetic to the idea that LLMs facilitate the creation of software that people previously weren't willing to pay for, but then kind of by definition that's not going to have a big topline economic impact.
Why did you add that "weren't willing to pay for" condition?<p>Most of the software I replaced was software I was paying for (iStat Menus, Wispr Flow, Synology/Unraid). That I was paying for a project I could trivially take on with AI was one of the main incentives to do it.
It's on the people pushing AI as the panacea that has changed things to show workings. Not someone saying "I've not seen evidence of it". Otherwise it's "vibes" as you put it.
Here's an example: <a href="https://eudaimonia-project.netlify.app/" rel="nofollow">https://eudaimonia-project.netlify.app/</a><p>I'm happy to sell it to you, though it is also free. I guided Claude to write this in three weeks, after never having written a line of JavaScript or set up a server before. I'm sure a better JavaScript programmer than I could do this in three weeks, but there's no way I could. I just had a cool idea for making advertising a force for good, and now I have a working version in beta.<p>I'd say it is better software, but better is doing a lot of heavy lifting there. Claude's <i>execution</i> is average and always will be, that's a function of being a prediction engine. But I genuinely think the <i>idea</i> is better than how advertising works today, and this product would not exist at all if I had to write it myself. And I'm someone who has written code before, enough that I was probably a somewhat early adopter to this whole thing. Multiply that by all the people whose ideas get to live now, and I'm sure some ideas will prove to be better even with average execution. Like an llm, that's a function of statistics.
And now you have no idea how any of the code works<p>AI writes bad software by virtue of it being written by the AI, not you. No actual team member understands what's going on with the code. You can't interrogate the AI for its decision making. It doesn't understand the architecture its built. There's nobody you can ask about why anything is built the way it is - it just exists<p>Its interesting watching people forget that the #1 most important thing is developers who understand a codebase thoroughly. Institutional knowledge is absolutely key to maintaining a codebase, and making good decisions in the long term<p>Its always been possible to trade long term productivity for short term gains like this. But now you simply have no idea what's going on in your code, which is an absolute nightmare for long term productivity
You can read as much or as little of the code as you want.<p>The status quo was that I have no better understanding of code I haven't touched in a year, or code built by other people. Now I have the option to query the code with AI to bootstrap my understanding to exactly the level necessary.<p>But you're wrong on every claim about LLM capabilities. You can ask the AI exactly why it decided on a given design. You can ask it what the best options were and why it chose that option. You can ask it for the trade-offs.<p>In fact, this should be part of your Plan feedback loop before you move to Implementation.
You can ask the AI why, but its answer doesn't come from any kind of genuine reasoning. It doesn't know why it did anything, because it doesn't exist as a sentient being. It just makes something up that sounds good<p>If you choose to take AI reasoning at face value, you're choosing to accept pretty strong technical debt
My own observation is that the initial boost to productivity results in massive crippling technical debt.
Having the productivity "drop through the floor" is a bit hyperbolic, no? Humans are still reviewing the PRs before code merge at least at my company (for the most part, for now).
Ironically, abstraction bloat eats away any infra gains. We trade more compute to allow people less in tune with the machine to get things done, usually at the cost of the implementation being eh... Suboptimal, shall we say.
> There are many reasons for the lag in productivity gain but it certainly will come.<p>Predictions without a deadline are unfalsifiable.
My unfounded hunch for the computing bit is that home computers became more and more commonplace in the home as we approached the 21st century.<p>A Commodore 64 was a cool gadget, but “the family computer” became a device that commoditized the productivity. The opportunity cost of applying a computer to try something new went to near zero.<p>It might have been harder for someone to improve the productivity of an old factory in Shreveport, Louisiana with a computer than it was for the upstarts at id to make Doom.
In my org I get far more done than ever, but I also find it more exhausting.<p>Because I can get so much done, I've lost my sense for what's enough. And if I can squeeze out a bit more relatively easily, why wouldn't I? When do I hit the brakes?<p>There are some tasks where LLMs are not all that helpful, and I find myself kind of savoring those tasks.<p>I'm surprised you don't notice a difference. Where I work it has been transformative. Perhaps it's because we're relatively small and scrappy, so the change in pace is easier with less organizational inertia. We've dramatically changed processes and increased outputs without a loss in quality. For less experienced programmers who are more interested in simple scripts for processing data, their outputs are actually far better, and they're learning faster because the Claude Code UI exposes them to so many techniques in the shell. I now see people using bash pipes for basic operations who wouldn't have known a thing about bash a couple years ago. The other day a couple less-technical people came to me to learn about what tests are. They never would have been motivated to learn this before. It's really cool.<p>It doesn't reduce work at all, though. We're an under-funded NGO with high ambition. These changes allow us to do more with the same funding. Hopefully it allows us to get more funding, too. I can't see it leading to anyone being let go; we need every brain we can get.
My employer is pretty advanced in its use of these tools for development and it’s absolutely accelerated everything we do to the point we are exhausting roadmaps for six months in a few weeks. However I think very few companies are operating like this yet. It takes time for tools and techniques to make it out and Claude code alone isn’t enough. They are basically planning to let go of most of the product managers and Eng managers, and I expect they’re measuring who is using the AI tools most effectively and everyone else will be let go, likely before years end. Unlike prior iterations I saw at Salesforce this time I am convinced they’re actually going to do it and pull it off. This is the biggest change I’ve seen in my 35 year career, and I have to say I’m pretty excited to be going through it even though the collateral damage will be immense to peoples lives. I plan to retire after this as well, I think this part is sort of interesting but I can see clearly what comes next is not.
I’m observing very similar trends at a startup I’m at. Unfortunately I’m not ready to retire yet.
Why are you excited for this? They’re not going to give YOU those peoples’ salaries. You will get none of it. In fact, it will drag your salary through the floor because of all the available talent.
I’m excited as a computer scientist to see it happening in my life time. I am not excited for the consequences once it’s played out. Hence my comment about retiring, and empathy for everyone who is still around once I do. I never got into this for the money - when I started engineers made about as much as accountants. It’s only post 1997 or so that it became “cool” and well paid. I am doing this because I love technology and what it can do and the science of computing. So in that regard it’s an amazing time to be here. But I am also sad to see the black box cover the beauty of it all.
I'm very confused about this. Salary is only one portion of your total compensation. The vast majority of tech companies offer equity in a company. The two ways to increase the FMV of your equity is: increase your equity stake or increase the value of the total equity available. Hitting the same goals with fewer people means your run rate is lower, which increases the value of your equity (the FMV prices in lower COGS for the same revenue.) Also, keeping on staff often means you want to offer them increased equity stakes as an employment package. Letting staff go means more of that available equity pool is available to distribute to remaining employees.<p>We aren't fungible workers in a low skill industry. And if you find yourself working in a tech company without equity: just don't, leave. Either find a new tech company or do something else altogether.
The dev team is <i>committing</i> more than they used to. A lot, in fact, judging from the logs. But it's not showing up as a faster cadence of getting me software to administer. Again, maybe that will change.
I think they feel more productive but aren't actually.
In my experience it is now twice the amount of merge requests as a follow-up appears to correct any bugs no one reviewed in the first merge request.
I’m at a big tech company. They proudly stated more productivity measures in commits (already nonsense). 47% more commits, 17% less time per commit. Meaning 128% more time spent coding. Burning us out and acting like the AI slop is “unlocking” productivity.<p>There’s some neat stuff, don’t get me wrong. But every additional tool so far has started strong but then always falls over. Always.<p>Right now there’s this “orchestrator” nonsense. Cool in principle, but as someone who made scripts to automate with all the time before it’s not impressive. Spent $200 to automate doing some bug finding and fixing. It found and fixed the easy stuff (still pretty neat), and then “partially verified” it fixed the other stuff.<p>The “partial verification” was it justifying why it was okay it was broken.<p>The company has mandated we use this technology. I have an “AI Native” rating. We’re being told to put out at least 28 commits a month. It’s nonsense.<p>They’re letting me play with an expensive, super-high-level, probabilistic language. So I’m having a lot of fun. But I’m not going to lie, I’m very disappointed. Got this job a year ago. 12 years programming experience. First big tech job. Was hoping to learn a lot. Know my use of data to prioritize work could be better. Was sold on their use of data. I’m sure some teams here use data really well, but I’m just not impressed.<p>And I’m not even getting into the people gaming the metrics to look good while actually making more work for everyone else.
Lol its gonna take longer than it should for this to play out.<p>Sunk cost fallacy is very real, for all involved. Especially the model producers and their investors.<p>Sunk cost fallacy is also real for dev's who are now giving up how they used to work - they've made a sunk investment in learning to use LLMs etc. Hence the 'there's no going back' comments that crop up on here.<p>As I said in this thread - anyone who can think straight - Im referring to those who adhere to fundamental economic principles - can see what's going on from a mile away.
Management is just stupid sometimes. We had a similar metric at my last company and my manager's response was "well how <i>else</i> are we supposed to measure productivity?", and that was supposed to be a legitimate answer.
I don't doubt your sincerity. But this represents an absolutely bonkers disparity compared to the reality I'm experiencing.<p>I'm not sure what to say. It's like someone claiming that automobiles don't improve personal mobility. There are a lot of logical reasons to be against the mass adoption of automobiles, but "lack of effectiveness as a form of personal mobility" is not one of them.<p>Hearing things like this does give me a little hope though, as I think it means the total collapse of the software engineering industry is probably still a few years away, if so many companies are still so far behind the curve.
> It's like someone claiming that automobiles don't improve personal mobility.<p>I prefer walking or cycling and often walk about 8km a day around town, for both mobility and exercise. (Other people's) automobiles make my experience worse, not better.<p>I'm sure there's an analogy somewhere.<p>(Sure, automobiles improve the <i>speed</i> of mobility, if that's the only thing you care about...)
I feel that it differs a lot between companies. It seems like corporate are having less an impact for now, as external innovation needs tailoring to adapt to its needs (e.g a security solution that needs 3 month projects to be tailored to the company tech stack), whereas startups and smaller firms see the most of the impact so far.
A tool with a mediocre level of skill in everything looks mediocre when the backdrop is our own area of expertise and game changing when the backdrop is an unfamiliar one. But I suspect the real game changer will be that everyone is suddenly a polymath.
This is not a good thing. If you’re not being exposed and skilling up already, you’re likely to be in the camp that is washed away.<p>If you can’t be exposed to it in your day job, start using Claude opus in the evening so you know what’s coming.
So far I have not seen much skill gain from using LLM extensively.<p>Maybe I will be replaced by matrix multiplication in my job, but if I need to use LLM at some point I expect little benefit from starting now.<p>Yes, I tried to use Claude Code two months ago. It was scary, but not useful.
“Not useful” —- one of those moments where you have to be able to adjust your views in the face of new evidence. Humans are so wedded to their beliefs that it can be agonising to let go. I have nothing but respect for people who admit they were wrong, though. I remained a skeptic for a long time, but 4.5 was enough to convince me to adopt for production code.
So far I updated from "meh, not useful" to "scary, not useful".<p>If it would be useful I would continue to use it, but at this point I would not use even if it would be free, not proprietary and not funding replacing me.
> so far AI has had almost no impact on my job.<p>Are you hiring?
My company has been hiring a ton over the last year or so. Jobs are out there
My friend used to say that, and he got quietly fired and outsourced because now someone in India can use ChatGPT to produce similar code, lol.<p>IMO AI will make 70-80% job obsolete for sure.
But, as I said above, I don't produce code; I administer it (administrate? whichever it is).
>now someone in India can use ChatGPT to produce similar code,<p>lol, that sounds like a disaster for the codebase.
I will personally say right now... its not gonna change lol.<p>People who actually know how to think can see it a mile away.
Build a new feature. If you aren't bogged down in bureaucracy it will happen much faster.
I dont use LLMs much. When I do, the experience always feels like search 2.0. Information at your fingertips. But you need to know exactly what you're looking for to get exactly what you need. The more complicated the problem, the more fractal / divergent outcomes there are. (Im forming the opinion that this is going to be the real limitations of LLMs).<p>I recently used copilot.com to help solve a tricky problem for me (which uses GPT 5.1):<p><pre><code> I have an arbitrary width rectangle that needs to be broken into smaller
random width rectangles (maintaining depth) within a given min/max range.
</code></pre>
The first solution merged the remainder (if less than min) into the last rectangle created (regardless if it exceeded the max).<p>So I poked the machine.<p>The next result used dynamic programming and generated every possible output combination. With a sufficiently large (yet small) rectangle, this is a factorial explosion and stalled the software.<p>So I poked the machine.<p><i>I</i> realized this problem was essentially finding the distinct multisets of numbers that sum to some value. The next result used dynamic programming and only calculated the distinct sets (order is ignored). That way I could choose a random width from the set and then remove that value. (The LLM did not suggest this). However, even this was slow with a large enough rectangle.<p>So <i>I</i> poked my brain.<p><i>I</i> realized I could start off with a greedy solution: Choose a random width within range, subtract from remaining width. Once remaining width is small enough, use dynamic programming. Then I had to handle the edges cases (no sets, when it's okay to break the rules.. etc)<p>So the LLMs are useful, but this took 2-3 hours IIRC (thinking, implementation, testing in an environment). Pretty sure I would have landed on a solution within the same time frame. Probably greedy with back tracking to force-fit the output.
I just tested this with Claude Code and Opus 4.6, with the following prompt:<p>"I have an arbitrary width rectangle that needs to be broken into smaller random width rectangles (maintaining depth) within a given min/max range. The solution needs to be highly performant from an algorithmic standpoint, well-tested using TDD and Red/Green testing, written in python, and not have any subtle errors."<p>It got the answer you ended up with (if I'm understanding you correctly) the first time in just over 2 minutes of working, and included a solid test suite examining edge cases and with input validation.
How can we verify if you dont post the code?<p>I appreciate you testing, even though it's not a great comparison:<p>- My feedback cycle of LLM prompting forced me to be more explicit with each call, which benefited your prompt since I gave you exactly what to look for with fewer nuances.<p>- Maybe GPT 5.1 is old or kneecapped for newer versions of GPT<p>- Maybe Opus/Claud is just a way better model :P<p>Please post the code!<p>Edit: Regarding "exactly what to look for", when solving a new problem, rarely is all the nuance available for the first iteration.
> I don't use LLMs much<p>Sorry to be so blunt, but it's not surprising that you aren't able to get much value from these tools, considering you don't use them much.<p>Getting value from LLMs / agents is a skill like any other. If you don't practice it deliberately, you will likely be bad at it. It would be a mistake to confuse lack of personal skill for lack of tool capability. But I see people make this mistake all the time.
Most of these are new features, but then they have to integrate with the existing software so it's not really greenfield. (Not to mention that our clients aren't getting any faster at approving new features, either.)
Its this kind of thinking that tells me people cant be trusted with their comments on here re. "Omg I can produce code faster and it'll do this and that".<p>No simply 'producing a feature' aint it bud. That's one piece of the puzzle.
I've taken to calling LLMs processors. A "Hello World" in assembly is about 20 lines and on par with most unskilled prompting. It took a while to get from there to Rust, or Firefox, or 1T parameter transformers running on powerful vector processors. We're a notch past Hello World with this processor.<p>The specific way it applies to your specific situation, if it exists, either hasn't been found or hasn't made its way to you. It really is early days.
From my experience as a software engineer, doubling my productivity hasn’t reduced my workload. My output per hour has gone up, but expectations and requirements have gone up just as fast. Software development is effectively endless work, and AI has mostly compressed timelines rather than reduced total demand.
There's a famous quote by a cyclist, "It never gets easier, you just go faster"
It is not going to reduce your workload. It is going to remove one of your co-workers.
This seems unlikely. My company is in competition with a number of other startups. If AI removes one of my co-workers, our competitors will keep the co-worker and out-compete us.
> If AI removes one of my co-workers, our competitors will keep the co-worker and out-compete us.<p>This assumes that the companies' business growth is a function of the amount of code written, but that would not make much sense for a software company.<p>Many companies (including mine) are building our product with an engineering team 1/4 the size of what would have been required a few years ago. The whole idea is that we can build the machine to scale our business with far fewer workers.
How many companies have you worked at in the past where the backlog dried up and the engineering team sat around doing nothing?<p>Even in companies that are no longer growing I've always seen the roadmap only ever get larger (at that point you get desperate to try to catch back up, or expand into new markets, while also laying people off to cut costs).<p>Will we finally out-write the backlog of ideas to try and of feature requests? Or will the market get more fragmented as more smaller competitors can carve out different niches in different markets, each with more-complex offerings than they could've offered 5 years ago?
> This seems unlikely<p>This is already happening. Fewer people are getting hired. Companies are quietly (sometimes not, like Block) letting people go. At a personal level all the leaders in my company are sounding the “catch up or you’ll be left behind” alarm. People are going to be let go at an accelerated pace in the future (1-3 years).
I don’t think that addresses my point. I understand a lot of companies are firing under the guise of AI, but it’s unclear to me whether AI is actually driving this - especially when the article we are both responding to says:<p>> We find no systematic increase in unemployment for highly exposed workers since late 2022
It depends on the "shape" of the company. Larger companies have a lot more of what I call "Conway Overhead", basically a mix of legit coordination overhead and bureaucracy. Startups by necessity have a lot less of that, and so are better "shaped" to fully harness AI.
> This seems unlikely.<p>It is absolutely likely. The hiring market for juniors is fucked atm.
That's not necessarily a result of AI, you also have to consider the broader economic environment. I mean, it was also difficult to get a job as a graduate in 2008, whereas it's typically been easier to get a job when credit is cheap.
It sure was, but as far as I'm aware, 2026 isn't in the middle of a generation-scale economic collapse.<p>(And if it is, what is the cause?)
Isn't it, for something like 70-80% of families? Just in slow-motion?<p>How long have we been hearing about crushing affordability problems for property? And how long ago did that start moving into essentials? The COVID-era bullwhip-effect inflation waves triggered a lot of price ratcheting that has slowed but never really reversed. Asset prices are doing great, as people with money continue to need somewhere to put it, and have been <i>very</i> effective at capturing greater and greater shares of productivity increases. But how's the average waiter, cleaning-business sole-proprietor, uber driver, schoolteacher, or pet supply shopowner doing? How's their debt load trending? How's their savings trending?
There’s a difference between a collapse and a slowdown. We don’t need a collapse for hiring to slow down [1,2]. I think we’re finally just seeing the maturation of software development. Software is increasingly a commodity, so maybe the era of crazy growth and hiring is over. I don’t think that we need AI to explain this either, although possibly AI will simply commodify more kinds of software.<p>[1] <a href="https://www.npr.org/2026/02/12/nx-s1-5711455/revised-labor-department-figures-show-hiring-in-2025-was-lower-than-reported" rel="nofollow">https://www.npr.org/2026/02/12/nx-s1-5711455/revised-labor-d...</a><p>[2] <a href="https://www.marketplace.org/story/2025/12/18/expect-more-of-a-low-fire-low-hire-job-market-in-2026" rel="nofollow">https://www.marketplace.org/story/2025/12/18/expect-more-of-...</a>
FAANG realizing that they can't make infinite money by expanding into every possible market while paying FAANG salaries for low-scale-CRUD-prototyping roles has a lot to do with this, and that started a bit earlier than the AI wave.<p>Lots going on right now in the market, but IMO that retreat is the biggest one still.<p>Many companies were basically on a path of infinite hiring between ~2011 and ~2022 until the rapid COVID-era whiplash really drove home "maybe we've been overhiring" and caused the reaction and slowdown that many had been predicting annually since, oh, 2015.
Erm its been fucked for many years across many professions, it was just less so for software engineering in particular. Now entry into the S-E profession is taking a hit.<p>Also dont forget theres only so many viable revenue-generating and cost-saving projects to take. And said above - overhiring in COVID.
Because of overhiring during the post-COVID free money glitch, not because of AI.
Aren't we both responding to an article which says:<p>> We find no systematic increase in unemployment for highly exposed workers since late 2022
It was fucked before AI became "mainstream" too. Companies overhired during and after covid.
There's definitely tone deaf statements from managers/leaders like "AI will allow us to do more with less headcount!" As if the end worker is supposed to be excited about that, knuckleheads, lol.
In a bear market in a bloated company, maybe. We’re still actively hiring at my startup, even with going all-in on AI across the company. My PM is currently shipping major features (with my review) faster and with higher-quality code than any engineer did last year.
>My PM is currently shipping major features (with my review) faster and with higher-quality code than any engineer did last year<p>That's... not a good look for your engineers?
It’s hard to compare, honestly. Last year, my PM didn’t have the AI tools to do any of this, and engineers were spread thin. Now, the PM (with a specialized Claude Code environment) has the enthusiasm of a new software engineer and the product instincts of a senior PM.
> In a bear market in a bloated company, maybe<p>Then any company that was staffed at levels needed prior to the arrival of current-level LLM coding assistants is bloated.<p>If the company was person-hour starved before, a significant amount of that demand is being satisfied by LLMs now.<p>It all depends on where the company is in the arc of its technology and business development, and where it was when powerful coding agents became viable.
Or just make time for more Very Important Meetings.
This - I can't think of any place I've ever worked where development ever outpaced backlog and tech debt.
See: <a href="https://en.wikipedia.org/wiki/Jevons_paradox" rel="nofollow">https://en.wikipedia.org/wiki/Jevons_paradox</a>
At the risk of being the person who says, "it's capitalism," (I know I know).... When making profit is the dominant intent of a company, a worker doing something faster doesn't lead to the worker doing less. It leads to the worker producing more in the same time. If doing more yields too much of the thing produced for the market to handle, the company either A. creates more need for the more produced (fabricate necessity), or B. creates a new need for a new thing, and a new thing for you to produce. There's no getting off the wheel for the worker in capitalism.
The goal has always and will always be to complete as much as possible in the time allotted.
That’s the economy in general. Labor saving innovations increase productivity but do not usually reduce work very much, though they can shift it around pretty dramatically. There are game theoretic reasons for this, as well as phenomena like the hedonic treadmill.
I'm working on a project right now, that is <i>heavily</i> informed by AI. I wouldn't even try it, if I didn't have the help. It's a big job.<p>However, I can't imagine vibe-coders actually shipping anything.<p>I <i>really</i> have to ride herd on the output from the LLM. Sometimes, the error is PEBCAK, because I erred, when I prompted, and that can lead to very subtle issues.<p>I no longer review every line, but I also have not yet gotten to the point, where I can just "trust" the LLM. I assume there's going to be problems, and haven't been disappointed, yet. The good news is, the LLM is pretty good at figuring out where we messed up.<p>I'm afraid to turn on SwiftLint. The LLM code is ... <i>prolix</i> ...<p>All that said, it has <i>enormously</i> accelerated the project. I've been working on a rewrite (server and native client) that took a couple of years to write, the first time, and it's only been a month. I'm more than half done, already.<p>To be fair, the slow part is still ahead. I can work alone (at high speed) on the backend and communication stuff, but once the rest of the team (especially <i>shudder</i> the graphic designer) gets on board, things are going to slow to a crawl.
> However, I can't imagine vibe-coders actually shipping anything.<p>I'm a vibe-coder, and I've shipped lots! The key is to vibe-code apps that has a single user (me). Haven't coded anything for 15 years prior to January too.
I suspect we have different definitions of “ship.”<p>I am usually my principal customer, but I tend to release publicly.
So your the dev who wrote Tea then huh
>> I no longer review every line, but I also have not yet gotten to the point, where I can just "trust" the LLM.<p>Same here. This is also why I haven't been able to switch to Claude Code, despite trying to multiple times. I feel like its mode of operation is much more "just trust to generated code" than Cursor, which let's you review and accept/reject diffs with a very obvious and easy to use UX.
Most of the folks I work with who uninstalled Cursor in favor of Claude Code switched back to VSCode for reviewing stuff before pushing PRs. Which... doesn't actually feel like a big change from just using Cursor, personally. I tried Claude Code recently, but like you preferred the Cursor integration.<p>I don't have the bandwidth to juggle four independent things being worked on by agents in parallel so the single-IDE "bottleneck" is not slowing me down. That seems to work a lot better for heavy-boilerplate or heavy-greenfield stuff.<p>I <i>am</i> curious about if we refactored our codebase the right way, would more small/isolatable subtasks be parallelizable with lower cognitive load? But I haven't found it yet.
Is there a reason you don’t use a hook to make all code pass your linters before you look at it?
I’m probably gonna do that (I use SwiftLint[0] -Note: I no longer use CocoaPods anything, these days. I wrote that, years ago.), but I tend to be <i>quite</i> strict, and didn’t want to be constantly interrupting myself, polishing turds. It was really kind of a joke.<p>I haven’t turned it on, yet, because of the velocity of the work, but I think I’ve found my stride.<p>[0] <a href="https://littlegreenviper.com/swiftlint/" rel="nofollow">https://littlegreenviper.com/swiftlint/</a>
I don't think there's been much of an impact, really. Those who know how to use AI just got tangentially more productive (because why would you reveal your fake 10x productivity boost so your boss hands you 10x more tasks to finish?), and those w/o AI knowledge stayed the way they were.<p>The real impact is for indie-devs or freelancers but that usually doesn't account for much of the GDP.
Work is freezing hiring and upping spending on tokens for everyone.<p>Don't know if this is effective and I don't think management knows either, but it's what they're doing
> Work is freezing hiring and upping spending on tokens for everyone.<p>Doesn't mean the two are related.<p>Is AI just the excuse? We've got tariffs, war, uncertainty and other drama non stop.
>Is AI just the excuse?<p>When you're unemployed, it doesn't matter. When executives cargo cult, it doesn't matter.
It's what they're telling us
Of course they are.<p>Management often has a perverse short-term incentive to make labor feel insecure. It’s a quick way to make people feel insecure and work harder ... for a while.<p>Also, “AI makes us more productive so we can cut our labor costs” sounds so much better to investors than some variation of “layoffs because we fucked up / business is down / etc”
You should look into the concepts of skepticism, materialism, and cynicism. Maybe don't trust the leadership of where you work, the leadership that sees you as a number and not a human.
Do you believe everything management tells you, whether you’re internal or external?
Which story sends a more positive signal to shareholders?<p>"We've frozen hiring because our growth potential is tapped out."<p>"We've frozen hiring because AI can replace employees."
If everyone was 10x productive then we would have had native Claude Code app for each platform.<p>Instead they are using Electron and calling it a day. Very ironic isn't it? If AI is so good then why don't we get native software from Anthropic?
Shift from quality to velocity started not yesterday but AI only accelerated this shift. Majority of comments here tell how fast Claude can generate code, not how good the result is. Electron is the prefect fit for move fast mindset. It is likely that Claude developers don’t see electron as problem at all.
I understand why they used Electron. But since AI is so amazing they should be using Claude to generate super optimized native software! They do claim that Claude is better than humans and AI is replacing programmers any day now.<p>So why aren't they using their own software to generate a linux optimized package for linux, a Swift software for MacOS and whatever windows uses.<p>That would be the best ad for AI. See, we use our own product!<p>But it doesn't happen.<p>So essentially by not generating custom binaries for every platform and using Electron they're doing one thing but saying something else. So maybe generating code isn't the #1 problem in the world!<p>Also I remember them saying that their engg write less code and they use Claude to write Claude. If Claude can be used to write Claude then why not use Claude to write OS specific binaries?!
I'd be curious to see the shift in numbers since December, 2025.
[dead]
I am not going to trust a single word from a company whose business <i>is</i> selling you AI products.
I also thought it was hilarious that they invented a brand new metric that (surprise) makes their product’s long term projection look really good (financially).
The problem it's not intended for plebs.<p>It just becomes a source of truth for media and corporate decision machines.
... and eyeing an IPO.
Wait a second, did they measure exposure from Claude logs and just assumed impact?<p>Let's say I sell snake oil and I survey every buyer, trying to convince everyone doctors won't be needed in the future.<p>First conclusion is that retired population seeks medical services the most (reality check - according to CDC most doctor visits are for infants).<p>Second conclusion is that because it's a snake oil, it heals all the problems and those people will never return to outdated healthcare system.
One of the more interesting takes I heard from a colleague, who’s in the marketing department, is that he uses the corporate approved LLM (Gemini) for “pretend work” or very basic tasks. At the same time he uses Claude on his personal account to seriously augment his job.<p>His rationale is he won’t let the company log his prompts and responses so they can’t build an agentic replacement for him. Corporate rules about shadow it be damned.<p>Only the paranoid survive I guess
the numbers they show are barely distinguishable from noise as far as I can interpret them.<p>For me, the impact is absolutely in hiring juniors. We basically just stopped considering it. There's almost no work a junior can do that now I would look at and think it isn't easier to hand off in some form (possibly different to what the junior would do) to an AI.<p>It's a bit illusory though. It was always the case that handing off work to a junior person was often more work than doing it yourself. It's an investment in the future to hire someone and get their productivity up to a point of net gain. As much as anything it's a pause while we reassess what the shape of expertise now looks like. I know what juniors did before is now less valuable than it used to be, but I don't know what the value proposition of the future looks like. So until we know, we pause and hold - and the efficiency gains from using AI currently are mostly being invested in that "hold" - they are keeping us viable from a workload perspective long enough to restructure work around AI. Once we do that, I think there will be a reset and hiring of juniors will kick back in.
Doesn't make sense to stop hiring juniors.<p>If AI increases productivity, and juniors are cheaper to hire, but is just as able to hand off tasks to ai as a senior, then it makes more sense to hire more juniors to get them working with an AI as soon as possible. This produces output faster, for which more revenue could be derived.<p>So the only limiting factor is the possibility of not deriving more revenue - which is not related to the AI issue, but broader, macroeconomic issue(s).
Juniors are not as capable of delegating to AI as seniors are. Delegation to AI requires code review, catching the AI when it doesn’t follow good engineering practices, and catching the AI in semantic mistakes due to the AIs lack of broader context. Those things are all hard for juniors.
Isn't that the point, to be able to learn that.<p>The craft changes with all these AI helpers, so the juniors have to also catch up/change with it. Or there won't b any seniors in due time.
It's really not an easy problem to solve.<p>You would hire someone with the expactation that they learn, but you also need to pay them. New hires always slow the team down. And currently you wouldn't even get much out of them, as you can delegate those tasks to AI.<p>Additionally you can not even be sure that the junior will learn or just throw stuff at AI. The amount of vibecoded Code I have to review at the moment from Seniors is stunning.<p>So yeah, the market needs Seniors, but there is basically no incentive for a company to hire a Junior at the moment. It's just easier and cheaper to pay a bit better than the market and hire Seniors then to train a Junior for years.
> but is just as able to hand off tasks to ai<p>I think this is the crux of it. Someone who doesn't know the right thing to do just isn't in a position to hand off anything. Accelerating their work will just make them do the wrong thing faster.
Based on my experience with using AI for development work, it feels like you really need to work with it instead of expecting it to do the work for you. Rather than type the code yourself by hand, you now need to explain the task very clearly, then review or test the generated code and then ask it to refactor and fix the issues you identify. This in itself is work that needs to be done, a different way of working compared to manual coding, but that doesn't mean any significant overall productivity gains are always guaranteed.
The endgame will be workers competing with networks of AIs that can solve business problems at all levels.<p>I'm curious how the system will maneuver itself to deprive workers of pay so that they can stay competitive with the ever-decreasing cost of AI.<p>Conversely, I'm curious how disruptors will find ways to provide workers with pay (perhaps through mutual aid networks, grants and alternative socioeconomic systems) so that they can use AI to produce the resources they need outside of the contracting labor market.
AI is coming for jobs—but the real risk isn’t where most people are looking.<p>The leading AI exposure indices (Anthropic, Eloundou et al.) focus on which jobs get automated. They treat low exposure as “safe.”<p>But the least exposed workers—cooks, roofers, dishwashers, construction laborers—are often in the worst jobs: low pay, high physical toll, short career spans, and little upward mobility. Safe from AI, but not from burnout or injury.<p>I built JQADI (Job Quality-Adjusted Displacement Index) to combine AI exposure with job quality. It surfaces three kinds of risk:<p>High AI exposure → classic displacement risk
Low AI, low quality → “trapped” workers in grinding, unsustainable jobs
Moderate AI, low quality → partial automation strips cognitive work and leaves physical drudgery (the “task residual” effect)<p>Findings: 83.5M workers are in low-AI, low-quality jobs. Customer service reps, data entry keyers, and medical records specialists sit at the intersection of high exposure and poor quality. Meanwhile, chief executives and lawyers are both low-exposure and high-quality.<p>The index uses ONET, BLS, and Anthropic exposure data. Code and methodology are open source. LINK <a href="https://github.com/quinndupont/JQADI" rel="nofollow">https://github.com/quinndupont/JQADI</a>
I mean, if that's true, then the governments around the world better damn well be prepared.<p>There's a ton of millennials (myself included) turning 40, that have been in this field since 2005 or earlier. It's all we know, and at this point we're getting too old to just "go do physical labor for minimum wage so AI can write code instead." I'm certainly too old to go back to school and try to pass the bar example to be a lawyer at 50+, and I have zero interest in any kind of people management whatsoever.<p>IMO Anthropic, OpenAI, Google, etc. should all be helping governments work toward a plan and lobbying for regulation on it instead of just charging full steam ahead "damn the consequences, those are someone else's problem."<p>It's going to obliterate what little is left of the middle class and leave a massive amount of unemployed middle aged tech workers with no where to go. What then? We either get ahead of the problem now (Outlook not so good), or we collapse into massive civil unrest and chaos.
Agreed. My point for spinning up this alternative metric is that the policy implications of the original suggest people should go out and get tough, dirty, dangerous jobs if they don't want to be displaced. But, there's a reason you don't see many 60 year old people in the trades.
What’s your reasoning for labeling lawyers as low-exposure?
My partner is a lawyer (prosecutor for a large city). The reason she is at low risk is simply because of the rate of adoption of AI tooling (or ANY tooling for that matter). IT in the public sector (particularly city government) is so much worse than I ever could have imagined before meeting my partner.<p>Our city just spent >$15MM on "case management software" that took 5 years to build by some fly-by-night outfit in California who won the contract, haphazardly bolted together MSFT Azure components, then vanished with zero support.<p>These teams can't in good faith freely adopt AI tooling into their workflow because they don't have the bandwidth to do it well, so they don't do it at all.
That's largely based on the original analysis and their methodology. "Responsibility" (only attributable to humans) is one reason, another is that judges probably don't want to speak with robots in court.
I know kids avoiding many high paying careers because of ai right now, and artists just giving up everywhere i look. Thanks, ai
Art should be done foremost because it's a passion for the artist. If you give up art just because you can no longer sell it, because you're being out competed by computer generated furry porn, then the world hasn't really lost anything of value.
That isn't how it works. Jimi Hendrix, the Beatles, and most other innovators were regular working artists long before they were influential. If you take away the means for artists to sell work that isn't boundary-pushing, they won't get enough practice and exposure to make influential work later.
that seems a bit harsh. if you’re livelihood has been from making some sort of visual art for however many years and work has even drying up bc of AI, just doing a sudden career pivot is pretty difficult.
True, it's harsh. But the way I see it, artists producing real art from their soul are going to keep at it even while corporate graphic designers and furry porn fanart illustrators are getting decimated. Society at large won't be culturally impoverished by this, so I'm viewing corporate graphic artists with the same mindset as buggy whip manufacturers. Technology coming for the careers of artists isn't even new, the invention of photography greatly diminished demand for painted portrays. Sad for those painters, but photography was then made into an art of it's own, and similarly I think creatively driven people will find new ways to exploit these new tools.
I know multiple devs who would have a very large productivity increase but instead choose to slow down their output on purpose and play video games instead.
I get it.
Productivity up by 10%. Happiness, life satisfaction and feeling of self-worth down by 20%.
I think it really depends what you're working on. I do some consulting and found it's not helping the C++ devs as much it's helping the html/js devs.
Finally someone that says it.
I think it’s a multi-variable problem as it handles languages differently. Also working in legacy code can be worse more often than not.
C++ dev here. It’s helping me just fine. Not as much as frontend folks but not far behind either.<p>It’s not quite at the place where LLMs can take over 100% coding, but give it a few more months.
We are rewriting our entire frontend from Webpack + Gatsby to Vite + React, we converted all static pages in one day using Claude Code.<p>We basically have ~40 components and 6 pages to go until complete rewrite, I am sure we will run into bumps in the road, but it's been crazy to watch.<p>We also added i18n (English + Spanish), ThemeProvider for white labeling solution, and WCAG 2A compliance, all in one shot.<p>If I went to a third party and asked them to rewrite just the static pages it would have been $200k and 3 months of work.
If people think Elite Overproduction (<a href="https://en.wikipedia.org/wiki/Elite_overproduction" rel="nofollow">https://en.wikipedia.org/wiki/Elite_overproduction</a>) is causing strife now, wait until tens of thousands of people with degrees get thrown out of work.
My day to day is even busier now with agents all over the place making code changes. The Security landscape is even more complex now overnight. The only negative impact I see is that there’s not much need for junior devs right now. The agent fills that role in a way. But we’ll have to backfill some way or another.
A possible outcome of AI: domestic technical employment goes up because the economics of outsourcing change. Domestic technical workers working with AI tools can replace outsourcing shops, eliminating time-shift issues, etc at similar or lower costs.
The problem with using unemployment as a metric is hiring is driving by perception. You're making an educated guess as to how many people you need in the future.<p>Anthropic can cause layoffs through pure marketing. People were crediting an Anthropic statement in causing a drop in IBM's stock value, which may genuinely lead to layoffs: <a href="https://finance.yahoo.com/news/ibm-stock-plunges-ai-threat-142900181.html" rel="nofollow">https://finance.yahoo.com/news/ibm-stock-plunges-ai-threat-1...</a><p>We'll probably have to wait for the hype to wear off to get a better idea, but that might take a long while.
Between 2004 and 2008 I did many things in computing as a company that offered my services, one of these was information gathering automation. It almost never immediately lead to decreases in employment. The systems had a to be in place for a while, people had to get used to them, people had to stop making common mistakes with them.<p>Then the 2008 crash happened and those people were gone in a blink of an eye and never replaced. The companies grew in staff after that, but it was in things like sales and marketing.
My speed shipping software increased but so did the demands of features by my company.
I don't really get this TBH.<p>Shipping speed never/is was the issue. Most companies are terrible at figuring out what exactly they should be allocating resources behind.<p>Speeding up does not solve the problem that most humans who are at the top of the hierarchy are poor thinkers. In fact it compounds it. More noise, nice.
Or worse. I’ve heard stories from friends where leadership expects huge boosts in productivity due to LLMs, and perceive anything but an order of magnitude boost as incompetence or a refusal to adapt.
To be fair, some of it might genuinely be refusal to adapt? If we go by HN comments, there definitely do seem to be at least some people who are letting their hangups prevent them from learning this tech.<p>> <i>It can't think, it just predicts likely tokens</i><p>> <i>I can't believe this industry I once cherished for rational professionalism has fallen for nondeterministism</i><p>> <i>Sorry, I'm just not going to participate in destroying the planet with these power hunger DCs</i><p>> <i>All this stuff actually costs 10x what a human developer costs but they're dumping the service at a low price to make us dependent.</i><p>> <i>It's a bubble, or a scam, in a year or two everything will go back to normal.</i><p>Tell me sentiments like these don't get bandied about by devs who want to keep doing things the way they know and like.
I just don't want to use it. It may or may not be a big deal, I lean toward it not being, but even if it is I would prefer not to use it, so I'm going to resist it as far as possible. I don't pretend it's about any higher principle than me not liking it. I'll wait to get put on a PIP and then I'll do exactly what they ask.
PMs can now also ship their half-baked requirements documents even faster thanks to the help of AI.
One thing I’m noticing in organizations is that AI tends to amplify judgment rather than replace it.<p>I think experienced people move faster because they can evaluate the output and redirect it, less experienced people often struggle because they don’t yet know what “good” looks like.<p>The interesting long-term question is how companies rebuild the environments where that judgment gets developed in the first place.
Anthropic should be outsourcing this kind of studies by providing data to non-affiliated researchers instead of doing the analysis themselves.
There looks to be some errors in the conversion from PDF -> web in this report. For example, the web version of Figure 7 has the legend colours reversed.
> There's suggestive evidence that hiring of young workers (ages 22–25) into exposed occupations has slowed — roughly a 14% drop in the job-finding rate<p>There goes my excuse of not finding a job in this market.
this keeps me up at night. i’m in a role that is essentially deployment management for LLMs at faang esque company. very little coding or need to code, mostly navigating guis, pipelines, and docker to get deployments updated with a new venting or model version or some patch
I'm an SDE with 1 YOE using AI tools heavily (doing "day's work" in ~2 hrs, perfect reviews). Spending most time on specs/review vs. raw coding. Worried I'm optimising short-term output over long-term skill development. Should I consider pivoting to AI/ML roles? Would love advice from anyone who's hired juniors in the current era.
Data point of 1: Having hired juniors as a startup founder, I need more generalists than AI/ML specialists. AI application work right now is basically standard software engineering - you’re finding clever ways to supply the right context to a model within certain constraints.<p>No one knows what’s going to happen in the future. Yes there already are fewer SWE jobs than before because of AI, and yes the days of companies hiring new grads in droves at $300k+ packages are likely over. IMO all you can really do is study what you’re interested in, learn it deeply, and do good work with cool people. If unsure, it’s possible to go back to what you were doing before if the new path doesn’t work out.
Has this been peer-reviewed?
Most of my recent interviews have been mostly people telling me that I am an idiot because I can't leetcode proficiently enough, so those jobs going away doesn't really effect me and at the same time makes sense. LLMs should be good with the leetcode classics that are the basis for rating software development productivity.
How is Anthropic getting this data? Are they running science experiments on people's chat history? (In the app, API or both?)
I'm not really concerned about the availability of SW dev jobs, but I am concerned about the quality of them. For many companies the velocity (and quality, much to my chagrin) of the code you can produce doesn't really matter. What matters more is whether or not you're building the right thing, and too often you're not. These companies also tend to keep more headcount than seems justified, I think because they are gambling that a few employees are going to do something awesome but they don't know which ones. As AI gets better what will these companies do? I don't think they will fire a bunch of SW devs. I think instead they will embrace the slop and just take more shots, and crazier shots. It doesn't just give us something to do, it also gives a bunch of PHBs something to do.
Much like LLM output, this seems convincing at first glance but it’s stating assumption as fact; That assumption being “when LLMs get better”. They don’t say what that ceiling is, but then go on to say “it can’t represent someone in court”. Why not? It can reason about more law and precedent than any human can right? As a society surely we want a fair and by the book justice system?<p>Look at GPT 5.4 and Opus, we’re clearly hitting diminishing returns already and these guys are pumping unsustainable amounts of money into them.<p>I’m bullish on AI, it’s been a net positive for me and my team. All I see here though is propaganda disguised as science to convince businesses to shrink their engineering budgets and redirect it to AI companies.<p>TL;DR: AI company says AI is amazing, more at 10.
I call BS on this as the ones displaced aren’t in the workforce anymore. I haven’t been able to work in over a year. Despite me applying to over 200 jobs a month.
You know you're having a real impact when you have to self-report on the impact you're having.
> Claude is extensively used for coding, Computer Programmers are at the top, with 75% coverage<p>I think there are some advantages to being first.<p>It's time to re-evaluate strategies if we've been operating under the assumption that this is going to be a bubble, or otherwise largely bullshit. It definitely works. Not everywhere all the time, but often enough to be "scary" now. Some of my prior dismissals like "text 2 sql will never work" are looking pale in the face today.
This is a pretty interesting report.<p>The TL;DR is that there is little measurable impact (and I'd personally add "yet").<p>To quote:<p>"We find no systematic increase in unemployment for highly exposed workers since late 2022, though we find suggestive evidence that hiring of younger workers has slowed in exposed occupations"<p>My belief based on personal experience is that in software engineering it wasn't until November/December 2025 that AI had enough impact to measurably accelerate delivery throughout the whole software development lifecycle.<p>I have doubts that this impact is measurable yet - there is a lag between hiring intention and impact on jobs, and outside Silicon Valley large scale hiring decisions are rarely made in a 3 month timeframe.<p>The most interesting part is the radar plot showing the lack of usage of AI in many industries where the capability is there!
> My belief based on personal experience is that in software engineering it wasn't until November/December 2025 that AI had enough impact to measurably accelerate delivery throughout the whole software development lifecycle.<p>Gemini 3 and Opus 4.6 were the "woah, they're actually useful now!" moment for me.<p>I keep saying to colleagues that it's like a rising tide. Initially the AIs were lapping around our ankles, now the level of capability is at waist height.<p>Many people have commented that 50% of developers think AI-generated code is "Great!" and 50% think its trash. That's a sign that AI code quality is that of the median developer. This will likely improve to 60%-40%, then 70%-30%, etc...
Never trust a statistic you haven't forged yourself.
This rhymes with another recent study from the Dallas Fed: <a href="https://www.dallasfed.org/research/economics/2026/0224" rel="nofollow">https://www.dallasfed.org/research/economics/2026/0224</a> - suggests AI is displacing younger workers but boosting experienced ones. This matches what we see discussed here, as well as the couple similar other studies we've seen discussed here.<p>Also, it seems to me the concept of "observed exposure" is analogous to OpenAI's concept of "capability overhang" - <a href="https://cdn.openai.com/pdf/openai-ending-the-capability-overhang.pdf" rel="nofollow">https://cdn.openai.com/pdf/openai-ending-the-capability-over...</a><p>I think the underlying reason is simply because companies are "shaped wrong" to absorb AI fully. I always harp on how there's a learning curve (and significant self-adaptation) to really use AI well. Companies face the same challenge.<p>Let's focus on software. By many estimates code-related activities are only 20 - 60%, maybe even as low as 11%, of software engineers' time (e.g. <a href="https://medium.com/@vikpoca/developers-spend-only-11-of-their-time-coding-what-3a53f65982df" rel="nofollow">https://medium.com/@vikpoca/developers-spend-only-11-of-thei...</a>) But consider where the rest of the time goes. Largely coordination overhead. Meetings etc. drain a lot of time (and more the more senior you get), and those are mostly getting a bunch of people across the company along the dependency web to align on technical directions and roadmaps.<p>I call this "Conway Overhead."<p>This is inevitable because the only way to scale cognitive work was to distribute it across a lot of people with narrow, specialized knowledge and domain ownership. It's effectively the overhead of distributed systems applied to organizations. Hence each team owned a couple of products / services / platforms / projects, with each member working on an even smaller part of it at a time. Coordination happened along the heirarchicy of the org chart because that is most efficient.<p>Now imagine, a single AI-assisted person competently owns everything a team used to own.<p>Suddenly the team at the leaf layer is reduced to 1 from about... 5? This instantly gets rid of a lot of overhead like daily standups, regular 1:1s and intra-team blockers. And inter-team coordination is reduced to a couple of devs hashing it out over Slack instead of meetings and tickets and timelines and backlog grooming and blockers.<p>So not only has the speed of coding increased, the amount of time spent coding has also gone up. The acceleration is super-linear.<p>But, this headcount reduction ripples up the org tree. This means the middle management layers, and the total headcount, are thinned out by the same factor that the bottom-most layer is!<p>And this focused only on the engineering aspect. Imagine the same dynamic playing out across departments when all kinds of adjacent roles are rolled up into the same person: product, design, reliability...<p>These are radical changes to workflows and organizations. However, at this stage we're simply shoe-horning AI into the old, now-obsolete ticket-driven way of doing things.<p>So of course AI has a "capability overhang" and is going to take time to have broad impact... but when it does, it's not going to be pretty.
What's interesting from a practical standpoint: the paper confirms what we're seeing in SME deployments – AI augments, not replaces. But the real productivity gain only kicks in when you redesign the <i>process</i> around the AI, not just bolt it on. Most small businesses skip that step entirely and then wonder why their 'AI tool' isn't delivering. The organizational restructuring is the hard part, not the technology. Anyone here seen teams actually get this right systematically?
Did you all read about the aws outage for 13hrs because their autonomous AI agent decided to delete everything and write from scratch?
I really hate to say it, but this article in particular needs a tldr. The author does a web recipe take. Don't put the actual factual info upfront and require parsing through everything to find anything important.<p>Kinda done with this.<p>If you have something important to say, say it up front and back it up with literature later.
[dead]
[dead]
[dead]
[flagged]
[flagged]
[flagged]
[flagged]
> that my spouse and her colleagues use AI A LOT for diagnosis and treatment plans<p>I hope they know what they're doing.
Just like anything else, you either think "that's definitely wrong" or "huh, I guess that's probably it." If its really serious, you have to pause and make of a judgement call of course.
There was a recent anecdote from the head of radiology, Mayo Clinic I believe, that went something like this:<p>- AI has allowed radiologists to review a much higher rate of x-rays<p>- The above has led to a dramatic increase in need for faster processing, more storage of scans etc<p>- which in turn led to needing a bigger IT department to manage all of the additional workload<p>There was a similar anecdote about the IRS where the claim is they went from having N accountants to having much fewer accountants but now they need N IT people to manage the new systems.
cigarettes don't cause cancer! -cigarette companies