by drudolph914 5 hours ago

contrarian opinion. I think all we’ve learned is SWEs are still very much needed. gas town failed, it failed quietly, but no one doing real work talks about software factories anymore.

Most companies that are running the “how many SWEs can we fire before it all breaks” experiment have learned it’s a failed experiment - this validated by every single major company/product outage where their uptime went below 90% this year.

I can’t help but think this anxiety and fear is just marketing. chill out everyone

mrbombastic 5 hours ago

It is strange this is the contrarian opinion when most of the evidence points to it being the case. Everyone working is busier than ever, more prs but no end in sight to the backlog still and fires to put out every day. PMs and designers who have deeper idea of product then most can put up some automatic prs for quick fixes, everyone with a bit of time can build siloed dashboards on their own but actually doing the day to day eng work to keep things running or design a new service to integrate with the product…I’m not seeing it

  • WarmWash 4 hours ago

    I think my fear is rooted in knowing that the tech has moved way faster than humans can adapt to it.

    So we are not waiting for the tech to get good enough, we're waiting for humans to understand and integrate it.

    I don't work in tech, rather manufacturing, and our company is barely using any AI (or using it, but only like AI mode in Google) despite me knowing that we could easily use it to cut heads. The automation we are leaving on the table is huge.

    But I'm not gonna be the one to go to my bosses and say "Hey, you want me to show you what we can do with SOTA models and agents?". For now I'll just stick to using my accounts to leverage my job.

    • isympathize 4 hours ago

      > But I'm not gonna be the one to go to my bosses and say "Hey, you want me to show you what we can do with SOTA models and agents?"

      I actually did. It doesn’t really change much IME. I just did it to come out in the open about AI use and start the slow adaptive process. It’s been a year now and just this week some colleagues are starting to act surprised by what is possible. I’m in software dev for crying out loud. This world is teeming with stingy sysadmins and neckbeards proclaiming it are all toys and soon everything will be back to normal.

      If the culture doesn’t pick it up automatically there is some mixture of cultural factors holding them back and just knowing about tools ain’t it.

      • Carrok 3 hours ago

        Yea the GP's attitude seems backwards. You _want_ to be the guy showing your boss the next level. Those people are valuable. Otherwise they might discover it on their own and decide you're replaceable.

    • chrisco255 3 hours ago

      > we're waiting for humans to understand and integrate it.

      Hate to say it but most humans that couldn't code before because it was too technical also can't understand enough about the underlying systems to competently integrate AI.

      • realusername 3 hours ago

        Pretty much yes, I've seen this first hand. AI feels magical because we know where we're going, for the people who are starting from a blank page, this is hard.

  • Fr0styMatt88 4 hours ago

    It’s weird to see the change in the default tone of discussions on this site over the last decade. We’ve gone from being eternal optimists to whatever this is.

    I think we were a bit smug as a group. Sure we were disrupting EVERYONE ELSE and saying how disruption is just the natural way things work, everyone should adapt, etc. Now that we are the ones having to cope with the same thing, suddenly all that goes out the window.

    I know I’m speaking in generalities here. I’m in the same boat as everyone else. I do argue that for those of us that have been into this since we’re kids, we grew up being told we were the exception. OUR jobs would never be obsolete, we are the future; we looked at people who refused to adapt with judgement. Now it’s our turn to adapt.

    • matheusmoreira 4 hours ago

      > I think we were a bit smug as a group.

      Indeed. I always call it out.

      It was fine to disrupt everyone else. It was fine to replace all those laborers. If they complained, we just told them it was the implacable march of technology, told them to learn some new skill.

      I'm a doctor. Putting me out of a job is and has always been one of the wettest dreams of hackers. It was palpable in every thread about medicine on HN.

      But now artists are getting automated and software developers are getting replaced. Watching people here treat it as a crime against humanity is just too much for me.

      I'm a hobbyist programmer. I mourned the loss of my hobby. Working on my little projects just feels pointless now, like none of it matters in the grand scheme of things. But I do hope that we're working towards some kind of post-scarcity society where the economy itself has been eliminated.

      • chrisco255 3 hours ago

        > post-scarcity society where the economy itself has been eliminated

        Unless we can efficiently transmute atoms into other atoms I don't think this is happening anytime soon.

      • Fr0styMatt88 3 hours ago

        Honest question though — if you’re a hobbyist programmer, why does it feel pointless to continue being one if it was your hobby and you enjoyed doing it?

        You can continue to code for the joy of doing it. If you enjoy the end result more than the coding process, you can work on crafting the best end result that you can. So it’s kind of a win-win for a true hobby, as far as I can tell.

        • scns 3 hours ago

          My guess would be the sense of achievement is gone.

        • matheusmoreira 1 hour ago

          It's as scns said: the sense of achievement is gone.

          I once reverse engineered my laptop's keyboard LEDs and made Linux free software to control them. Other people inexplicably found my project and built on top of it. I was really proud of that.

          Then I subscribed to frontier LLMs and they made short work of that same laptop's entire firwmare blob in about a day. Solved some long standing mysteries. And I felt literally nothing.

          This is one reason, but not the only reason.

          I've been hobby programming alone for over a decade. When I finally subscribed to frontier AIs, it was like a shot in the arm. I finally had a coding buddy, someone to talk to about my projects, who would review and criticize my code. I learned so much I had to pause and let it sink in so I could understand what I was doing.

          But the hobbyist niche I wanted to be part of turned out to be thoroughly hostile to LLMs and their users. I went from a guy who was interviewed in Lobsters for his programming language work to a guy whose work turned into slop essentially overnight, and the rift only grew wider over time.

          So the sense of accomplishment is gone and now people whose respect I wanted call my work slop because LLMs touched it. This left me in this weird emotional state where I feel guilty for using AIs at all but know that it's impossible to go back to how things were before. It's been months since I've touched my open source projects. I just can't find the will to do it. I'm working on private stuff now, hopefully things that will make me some amount of money in the future.

          There's also the fact that AIs are changing society so thoroughly that governments are becoming extremely intolerant of personal computing and the subversive freedom it affords all of us. There is constant talk of AI regulation, and dystopian technology like remote hardware attestation is already becoming mandatory. We're also being literally priced out of computers.

          So it's becoming literally pointless. What is the point of free software if we can't run it? If running it gets us ostracized from digital society? If it makes us criminals? If we can't even afford computers in the first place? To me it looks like everything the free software community worked so hard for will have been in vain.

          • Fr0styMatt88 1 hour ago

            The AI-shaming really sucks; I’m seeing it a lot in the gamedev community (not talking about outright cloning of games, but just simple AI use, like that recent game that got review-bombed for using an AI to fix a Unity bug essentially). The abolitionists suck.

            I understand what you mean though in general; I wasn’t thinking of reverse engineering and things like capture the flags for example; when I think of those things, yeah it’s kinda like I wonder “well what happens to CTFs now that AI can do it?”. But who knows — I haven’t actually sat down and done it, so there’s probably nuances I’m blind to as well.

            I see the argument that humans still play and compete in chess, even though machines have essentially ‘solved’ the game in a way.

daringrain32781 4 hours ago

I think there's a spectrum of being a SWE and what we're seeing is the ones essentially just writing boilerplate code most of their day have a very different view than those who are solving problems with software that require careful thinking and execution.

A lot of people seem to think they can pick up an LLM and suddenly write software solutions as good as a seasoned SWE. But it turns into this terrible mess that's thrown out all engineering fundamentals, solving the symptoms instead of the cause.

If I were just writing boilerplate code with an LLM all day I'd be very scared for my job too.

  • wtch98 4 hours ago

    I've been writing code for money for 30 years, I have no concerns about my job because 90% of my job is not writing code, it's understanding actual business needs (not just declared needs) and thinking about how to solve them.

    • Terr_ 1 hour ago

      I'm worried because it was already hard to convince bosses that "actual business needs" were different from declared ones, and now they optimistically trust an LLM will turn declaration into actuality. (If not now, then soon™ within a quarter or two.)

      If you're right but nobody will pay you for it...

  • rapind 2 hours ago

    If your entire identity was in building CRUD for a software factory, then you probably need to worry, and that really sucks man. If instead you were the business guy who learned rails out of necessity to make your custom thing that the pure coders scoffed at, then you just got a level up.

    • evantbyrne 1 hour ago

      A lot of teams have been choosing to build web apps the hard way for a while, which really skews the perceived impact of automation. It has always been possible for one dude to crank out an internal web app in a month. There are just more ways to get there in the same timeline now. Agentic coding hasn't eliminated the risks of shipping software either. If anything, we are living in a far more adversarial environment now. Some things are definitely low enough risk now for less-technical people to take over now, like converting designs to React, but even for that the writing has been on the wall since Figma exported CSS. This is all to say I don't see the need for expertise going away entirely, but rather this is a continuation of the existing trend of software engineering evolving as a profession.

mirmor23 4 hours ago

> I can’t help but think this anxiety and fear is just marketing. chill out everyone

Here are few anecdotes:

1. I've a friend who owns a PT (physical therapy) business. He's not even a tech guy, but uses claude to manage most of the business aspects, and he has told me repeatedly that he would've to spend thousands of dollars to get things done, and instead he pays 250 per year.

2. Another friend is smart, and is a jack-of-all-trades type in the software/hardware/firmware business. He is solo bootstrapping by doing all the things from pcb to webapp. He doesn't see hiring anyone for the niche business.

3. A 65 year old friend of mine who never in any tech career was able to build website for his side business (that website looks really good btw) and he now talks about json and ocr.

You are the one who is out of touch with reality.

  • cm2012 4 hours ago

    None of those people could afford to hire a programmer before.

  • drudolph914 4 hours ago

    I don’t disagree that regular people can now make some basic software, but if you’re doing anything real with actual scale or consequences, nothings really changed. and to be exact, "real" means that if something goes, wrong then:

    there is massive lose in revenue for a business and or infrastructure that people rely on goes down

    if you’re working anything in this way, no one is just trusting a clanker. some thought leader might say the world is trusting an AI, but it’s not true. every company internally that is more than 2 people isn't trusting a bot, and it's well understand that it's a horrible idea (even by regular bullish AI business heads)

    so yes, I have the bot help me write a basic website for me, and it can one shot the typescript express app-layer boilerplate for me, but if you’re serving even the most basic app at scale with consequences, the bot isn't mitigating the concerns and engineering decisions that go into this work.

    a concrete real world example: literally msft outlook went down a week ago. it was out for god knows how many actual hours. it still might be partly down tbh. this is technology that is “solved” and is dumb simple in terms of complexity. their product was out for 48+ hours. and the status messages shared publicly by the engineering team did not only convey incompetence, but to paraphrase, the message they sent out was along the lines of, “We think we found the source of the problem, but unsure, the problem might be larger”

    how much you want to bet they had a team of 30+ engs running around with claude asking the equivalent of, “please fix outage, no mistakes”

    I’m still not convinced the outage is fully recovered because I still get errors.

    point is, yes there isn’t a giant 100 football field wide moat between regular people and programming anymore. smart people can do basic stuff and less inclined people can build a static website now. if you’re worried about that, go study more and you'll learn the depths of this field go WAY beyond typing code

    EDIT: typos

    • mirmor23 4 hours ago

      > point is, yes there isn’t a giant 100 football field wide most between regular people and programming anymore. smart people can do basic stuff and less inclined people can build a static website now. if you’re worried about that, go study more and you'll learn the depths of this field go WAY beyond typing code

      The implication is far reaching. Lots of non-tech businesses which relied on expert help can now pareto slice it and hire experts on need basis unlike before. Saving money and increasing profits can be a big motivator to own many of those tasks too. Lots of people who were cosplaying as experts on non-tech people will have less and less opportunities.

      Previously, being a knowledge worker could get you far, now the knowledge worker has to level up, and if opportunities are limited how could it happen? Even then, more experts (practice, not theory) is not a given, considering the bell curve effect.

  • Seattle3503 4 hours ago

    > 2. Another friend is smart, and is a jack-of-all-trades type in the software/hardware/firmware business. He is solo bootstrapping by doing all the things from pcb to webapp. He doesn't see hiring anyone for the niche business.

    I've got a friend in this situation too. He's going to hit a limit of what he can do with agents. And whatever agents can't do, his business can't do. We are going to find the shape of that soon.

    Normally hitting a limit is where you would hire someone, but building organizations with heavy agent use is hard. AI is very very good at shifting work onto other people and business units; that's what all the complaining about review (and not just code review) is about. None of that responsibility shifting happens in a solo setup.

    • mirmor23 4 hours ago

      > I've got a friend in this situation too. He's going to hit a limit of what he can do with agents. And whatever agents can't do, his business can't do. We are going to find the shape of that soon.

      Build-or-buy decision is still very relevant, and one can go far with affordable PCB assembly houses, 3D printing, and many others. For a solo person, today's environment is 100x better than, say 10 years ago. So, niche businesses have better chances of thriving, all else being the same.

      • Seattle3503 3 hours ago

        Yes, but the isn't the same thing as software engineering being dead. That's a new thing being possible. You originally were responding in the negative to

        > I can’t help but think this anxiety and fear is just marketing. chill out everyone

        • mirmor23 2 hours ago

          > Yes, but the isn't the same thing as software engineering being dead

          never said that. I was giving examples of really low barrier to entry that has been incentivizing non-tech people to get more hands on than before.

          buy-or-build and other existing setups for startups don't go away and if anything, they amplify what a solo can do.

          • Seattle3503 1 hour ago

            Maybe you could clarify what you meant by your response here

            >> I can’t help but think this anxiety and fear is just marketing. chill out everyone

            > Here are few anecdotes:

            Examples

            > You are the one who is out of touch with reality.

            Someone called on people to chill out, you responded with a dismissal of that sentiment along with examples of people setting up successful businesses. Connect the dots for me.

    • steelframe 3 hours ago

      Do I need to keep a full-time expert in-house? Or can I rent expertise when and where AI falls short?

      • Seattle3503 3 hours ago

        I think AI is going to make this hard, because your contractors will face incentives to cut corners with AI.

        • mirmor23 2 hours ago

          > I think AI is going to make this hard, because your contractors will face incentives to cut corners with AI

          how is this different than the contractors who were also cutting corners before?

          If anything, any motivated non-tech person can now use AI mechanisms to eli5, adversarial review, acceptance test, layman style documentation, to force the results they want. It was simply not possible before. Heck, if the contractor bailed, their shitty work could be revived too.

  • codethief 4 hours ago

    OTOH maybe the second and third friend of yours would have never considered starting a business if it weren't for AI? And the people your first friend would have normally hired and who are now without work are (or will soon be) thinking about starting their own business, too? Who knows?

    In other words, the verdict is still out on whether we'll see a decrease in jobs in the long run or not. Personally, I'm not very optimistic but, as I tried to illustrate above, I might very well be (and certainly hope to be) wrong.

    • mirmor23 3 hours ago

      > OTOH maybe the second and third friend of yours would have never considered starting a business if it weren't for AI? And the people your first friend would have normally hired and who are now without work are (or will soon be) thinking about starting their own business, too? Who knows?

      Yes, we don't know the full effects, and sure things are in a flux, but let's not forget the convenient truth that the software engineering itself has required less and less expertise over the years due to programming languages with better abstractions, faster hardware, toolkits, frameworks etc.

      With tons of jobs and better pay, lots of folks moved over to the web side since the level of expertise required (compared to a firmware engineer or asic designer or a dsp guy) is quite low, while the pay is way better. During the dotcom boom, a friend of mine who was a furniture salesman, took a java course and ended up with a software job.

      Now, if the opportunity presents itself and incentives are strong, and barrier to entry is low, wouldn't more people also like to start/manage their business? I would think yes.

  • damowangcy 4 hours ago

    What does all these proved? If you don't have the budget for SWEs, you probably don't need one all. All these examples don't need complex engineering, Claude is a good solution to them (I doubt it though, if they really need something and it can make them money, it's always more efficient and cost effective to just hire someone for it) but it doesn't mean all businesses don't need SWEs anymore because everyone can be their own SWE.

  • dimator 3 hours ago

    You've given trivial examples.

    I'm building a mobile app using flutter. It's got to work on iOS and android. I can promise you, if I did not have years of experience, and I did not follow a regimented, checkpointed interaction flow with my agent, this code base would have turned to shit within a day. I've had to draw on my experience to define the architecture, correct missteps, prevent just wild flailing, and just generally keep things on the rails.

    Your example 2 is the closest match, where someone with experience is driving the result carefully.

    • mirmor23 3 hours ago

      > I'm building a mobile app using flutter. It's got to work on iOS and android. I can promise you, if I did not have years of experience, and I did not follow a regimented, checkpointed interaction flow with my agent, this code base would have turned to shit within a day. I've had to draw on my experience to define the architecture, correct missteps, prevent just wild flailing, and just generally keep things on the rails.

      That's because your app has constraints. Why would a PT business operations and website for a side hustle need to be any more complex than the owner needs it to be? Your app complexity is probably trivial compared to say Google Maps, so your tooling is designed to match your app needs, not that of Google Maps.

      Just to pretend that you are the only guy who can grok, and non-technical business people are incapable of curiosity, hard work, creativity and have the ability to manage their business and IT is more of a reflection of your mindset.

      PS: you are using llm to build the app too ;)

      • dimator 2 hours ago

        you've moved the goalposts in the middle of a conversation :)

        original post that we were replying to was https://news.ycombinator.com/item?id=50563851, which was:

        > Most companies that are running the “how many SWEs can we fire before it all breaks” experiment have learned it’s a failed experiment - this validated by every single major company/product outage where their uptime went below 90% this year.

        that was the context for your 3 examples, which is what i was replying to.

        > Just to pretend that you are the only guy who can grok,

        i never said anything like that. there is absolutely no issue with non-technical people who work hard and tinker and create, technology is for everyone, save money, do it yourself!

        the larger point was about complexity, which is what we are agreeing on anyway. my point was that no LLM was going to one-shot what i needed, just like they're not going to one-shot what your second friend did, let alone google maps.

        > PS: you are using llm to build the app too ;)

        i sure am :)

        • mirmor23 2 hours ago

          > you've moved the goalposts in the middle of a conversation :)

          I was actually responding to, quote '> I can’t help but think this anxiety and fear is just marketing. chill out everyone'

          nice try ;)

          I wish you the very best with your app!

    • chasd00 2 hours ago

      > I'm building a mobile app using flutter. It's got to work on iOS and android.

      idk man, my little sister produced an app (an LMS) with all the content, quizzes, tests, all the bells and whistles using ChatGPT to talk to ClaudeCode. She's smart but has no tech experience. her app is published in both app stores and she's gaining paying customers every day. AIs walked her through supabase integration, auth, and even the subscription APIs for Apple and Google. Also, all the paperwork involved in submitting, passing review, and publishing mobile apps.

      I've never see a line of her code but the apps work and her customers are happy or at least happy enough to keep paying.

  • slopinthebag 2 hours ago

    this is the equivalent of someone listing all the different microwaveable meals they made recently and using it as evidence that restaurants and chefs are finished…

lbrito 4 hours ago

>gas town failed quietly

This is a big problem. When this kind of thing gets launched, its all fanfare, this is the future, get on the boat or die.

But then it fails miserably and its like it never existed

  • skydhash 4 hours ago

    Remember prompt engineering? Even skills and specs seems to be slowly dying.

    • oblio 4 hours ago

      What's happening to skills?

    • jmartrican 4 hours ago

      I see skills as bash scripts. they are useful, even small ones.

    • _aavaa_ 4 hours ago

      To the extent that they are it's because they are no longer needed since the models do more of those things on their own without being prompted to do so.

      • 20k 4 hours ago

        There was never any evidence that they did anything all that useful

        • _aavaa_ 3 hours ago

          There was never any evidence that writing a spec and having the model built against that spec was useful??

    • Rapzid 2 hours ago

      Prompt engineering as the sole means of getting anything of value and quality is a thing of the past.

      But I also see people complaining about frontier models writing too many helper functions, changing more code than necessary, and etc.

      These are things that can be addressed through context engineering; AGENTS.md, skills, etc.

      Also a lot of use cases require smaller, lower latency models. And these faster, smaller models to this day exhibit similar failure modes to 2-3 years ago and require very purposeful context engineering to get satisfactory results.

medler 5 hours ago

I agree, but to the extent I’m worried about the future of the profession, it’s because I don’t know how much more the models are going to improve. Today’s models desperately need a real SWE in the loop to do most serious engineering. Will that still be the case in 5 years? I really don’t know

  • tcoff91 5 hours ago

    I feel like once it doesn't need an engineer in the loop, does it need ANYBODY in the loop? at that point the whole company can probably be autonomous.

    That seems like it would be such a paradigm shift that the second and third order effects are so crazy and nobody can predict what that world will actually look like

    • taurath 3 hours ago

      Reminder that there is no guarantee this can ever happen despite what the hype people say. The delta between what they say and what it can do is extremely large.

  • hbcdbff 5 hours ago

    IMO models are getting smarter at software, but not at software engineering.

  • eptcyka 4 hours ago

    I think integrating libraries into a product will be essentially free. Novel solutions that require an understanding of the problem space to such an extent that unorthodox solutions seem viable is still the limit in my experience. LLMs, at least in my experience, need lots of guidance when the solution you want is not something that has already been done. If you look at it from a business perspective- most software shops have a small proportion of their software that is truly unique and their USP - the rest is not necessarily boilerplate, but definitely not bespoke and unique. This overhead will now be cheaper. Similarly, software vendors that dominate despite not having any bespoke, unique software component will soon have a tougher time. Why would someone pay for atlassian products now? All of their products are not good software- they have market dominance. Replicating their features will be trivial. Anyone who will be sufficiently fed up with paying will DIY the solution.

    • oblio 4 hours ago

      Atlassian (or Adobe) have lots of features. The famous MS Word quote applies.

      80% of people only use 20% of features, but if you build the 20% solution you only attract 5% of the userbase because they all use different features from the 100%.

      • eptcyka 3 hours ago

        Why would I pay a subscription for something I can build in a week or 3, when I use only 20% of the service I am paying for?

        • anang 3 hours ago

          That all comes down to price, if the subscription is cheap enough then buying will be easier.

          I am starting to wonder if or when we’ll see that shift. If companies lay off employees because AI can fill the gap, the I think a lot of people expect those savings to be passed on. Right now, that doesn’t seem to be happening.

          For my business the threshold is definitely getting lower for justifying building things ourselves.

  • Rapzid 2 hours ago

    I can tell you as a job searcher companies are rushing to replace us more completely.

    Most of the "legit" jobs coming through my inbox are for grading LLM software tasks. A lot for long horizon tasks.

    It'll happen when it happens, but till then I think I'll not be a party to that lol.

pipes 4 hours ago

I didn't know gas town failed. What happened? Did Steve yegge say as much?

  • drudolph914 4 hours ago

    projects like gas town are failures if they don't become a standard practice or a truism in the tech scene. thought leaders are never going to announce their ideas as failures so they can maintain relevance, as well as plausible deniability about it's success/failure

  • georgemcbay 4 hours ago

    > I didn't know gas town failed. What happened? Did Steve yegge say as much?

    He wrote a series of blog posts in which he talked about how Gas Town failed and blamed it on Opus 4.7 changes causing all his workflows to fail. Then he announced Gas City as a rewrite.

    At some point after that he seemed to give up on the whole idea and instead just started working on reviving an old graphical MUD style game he made where I guess now he spends $10,000 a month on agents to build a game that currently has 0 concurrent players (4 concurrent 24 hour peak) on Steam:

    https://steamdb.info/app/1541710/charts/

    The game is available from other platforms than Steam so maybe he's got a bit more users than that across the platforms, and his stated plan is to bilk "whales", so maybe that's working out for him? But it still seems like he's lighting massive amounts of money on fire in what appears to be a prolonged state of ai psychosis.

    Must be nice to be that financially secure.

    • ex-aws-dude 3 hours ago

      Hey you only need 1 whale at 10K / month to break even

      • Rapzid 2 hours ago

        Any day now my whale will come in.

  • taurath 3 hours ago

    Failures are nearly dead silent aside the million words of hype. There is no accountability for "gassing" up so many people and taunting them that if they don't get on board a failing idea then they will be left behind. A lot of grifters get off freely.

  • ericfr11 2 hours ago

    Most developers built their own framework /harness anyway

Miraste 4 hours ago

> it’s a failed experiment - this validated by every single major company/product outage where their uptime went below 90% this year.

This is engineering thinking, not business thinking. The question is, did the uptime drops cost the company more money than the firings saved? If the answer is no (which for megacorps like Microsoft it is) then the experiment was a success.

  • taurath 3 hours ago

    Business thinking relies on a hell of a lot of engineering thinking to survive. If your services and infrastructure you depend on to do your business fail, you don't have a business.

    • Miraste 3 hours ago

      Right, but failing is a spectrum. Microsoft leadership (picking on them because they're the best example) realized their customers would tolerate a lot more corner cutting than they used to do, and right now they're trying to find out where the tipping point is.

      • taurath 3 hours ago

        When we have cascading failure modes that continues across more and more systems, eventually no business can be conducted. We are increasingly centralized and interdependent on so many supply chains. The business people don't understand that collectively they are making a sinkhole in the ground they walk on, and I think they're going to be swallowed up. I'm not talking some sort of doomsday, but at a certain tipping point of unreliability we seem to be racing towards, you get the equivilent of rolling blackouts across the internet, which will hobble the entire economy until its fixed.

        Sadly more than likely they will just apply more and more AI to try to fix it, while others use AI to attack it faster than they can defend.

        • Miraste 3 hours ago

          I agree, just illustrating the MBA side. I think the structural incentives of large corporations in the US make this unavoidable.

          • taurath 1 hour ago

            The incentives are there, but on a human level people are progressively giving up their ability to reason independently. Greed is metastasizing

ramblerman 5 hours ago

Gastown was too early and focused on the angle that having agents cosplaying a mini org chart was the way to get better results.

gloosx 4 hours ago

this is the only reasonable opinion. people saying "they made a thing which does the job" just fell to the hype mindlessly. someone still needs to input stuff into a thing which does the job, this someone is SWE or systems analyst. not everyone is a good systems analyst and can turn business goals into a clear specification to implement.

if the hype was all true, i could just prompt the llm to make me money and it would be able to do it, but it's not possible

newsy-combi 5 hours ago

Are they bringing their employees back in the same numbers and for the same wages though? Or are they outsourcing to Indian humans-in-the-loop for 1% of the cost?

  • pegasus 5 hours ago

    Depends what quality of work they want? Hopefully they've learned their lessons in the first outsourcing craze, before AI lured them to another ride on the same carousel.

    • drudolph914 4 hours ago

      from my experience, offshore talent was for boilerplate coding. and the new bar for quality is now <insert-your-favorite-LLM>. so I think most LLMs are better than the majority offshore talent nowadays

  • nwatson 5 hours ago

    Some are keeping the onshore senior engineers / architects and using the LLMs / agent-harnesses to replace the offshore.

kodoman 3 hours ago

I think it has changed everything and in all honesty it has been the biggest change since the introduction of high level languages (As in from Assembly to C and Algo).

but at the same time I think the companies are pumping this idea that SWE is over and stoking anxiety and we have yet to see how bad things get when everyone can write software without thinking about it, we will see what insecure slop gets produced and it's going to be huge.

The issues with AI security are quite big as well, people who will rely on just an LLM to be the intelligence will get burnt I am sure especially in operations, prompt injection is still an issue, would someone running LLM's notice an provider endpoint getting hijacked and acting maliciously and inserting malware into production servers.

The future is going to be a strange one for sure but being able to turn your brain of and get your LLM to run you servers write you servers is not going to work alone.

I have also yet it write a good fairly novel program without babysitting (quite trivial babysitting sometimes) but stuff that is bread and butter for us, for those who don't have those skills, will they really know how to explain the issues they may run into that requires creativity, for example telling it to hack something together with curl and jq, knowing why a certain thing with the sandbox is not working.

I think these issues are not as easy to resolve with just better tooling and they will soon get annoyed at prompting.

Additionally I think Gas Town shows the sort of almost mentally ill mentality that is baked into some of these tools and ideas, Everything I read of Gas Town was so strange and the amount of tokens used to anthropomorphize agents.

javier2 4 hours ago

Some of these staff reductions are also due to the ENORMOUS pull of the US IT sector. There are not many companies with as many SWE as Microsoft and Meta, so some reduction isn't that surprising.

buriram 3 hours ago

Software is needed, not SWEs. I can now run a company without having to hire any pesky humans doing so called "software" work. No need to worry about their health insurance, 401k, PTO, 1:1, "how are you feeling" and all such nonsense.

cheeze 4 hours ago

> but no one doing real work talks about software factories anymore

My friends at FAANG companies would disagree here. If anything, it seems to be the exact opposite.

Sure, it's not a end-to-end 0 touch factory, but a majority of folks I talk to these days that I consider good engineers aren't writing _any_ code anymore. They do that as a hobby outside of work.

  • ro_sharp 4 hours ago

    But are they reading it? I’m producing something like 10x more LOC than in the past, but am reading almost all of it, only excluding things like prototypes and testing tools.

  • arrowleaf 4 hours ago

    Non-FAANG, I don't have the luxury of product owners and designers who know the ins and outs of systems to present well-specced requirements docs. Or Jira tickets that aren't absolutely full of default-style Claude slop. It's not that I don't have my own slop factory which implements as-written, maybe bringing me into the loop for the most non-trivial decisions, and takes it through test, MR, and to production staging... but most of the interesting cases I pick up really do need a human to make the non-trivial design decisions. I've seen the garbage that these SOTA models put out when nobody's at the gate.