(po-ru.com)
I can totally identify with that feeling. It must be how professional mathematicians are feeling as well.
And also that on the other side of that grief is something new and potentially exciting, but very different than what we're all used to.
However, most code is corporate CRUD stuff. The world sort of runs on CRUD. But it's rarely particularly fun, or beautiful, or fulfilling. It tends to be pretty ugly because there's rarely time to make it beautiful. So I'm not losing much joy there.
Additionally, even with an agentic workflow, my team still thinks deeply about architecture issues. We believe that well structured code is better for bots and humans alike. So I am still exercising my engineering brain; I'm writing blueprints and supervising the bricklayers now instead of laying bricks.
Does that match your teams' experience at all? Sometimes I feel like folks on these threads are working in a parallel universe where the majority of coding tasks and somehow fulfilling and fun and not 99% tweaking reports, trying to implement arcane and contradictory business logic, or fighting with shitty and frustrating externalities. Maybe people found joy in those aspects. I guess there are all types of people in the world. I mean, I'm a sicko who loves writing SQL, something that most coders seem to hate.
I've been direct with my team about it being a legitimate
feeling of grief and the grieving process that comes along
with it. It's been helpful for folks to hear that they aren't
alone in that feeling.
I find that I struggle more with personal projects, as I tend to take things where I'm not even in the skirts of the learning curve, and consequently redo my work a few times.
I have no idea where you work but 90% of companies are not like that. In fact, it's the opposite: filled with awful management, terrible timelines and legacy code.
Engineering had always been a nuisance to them and getting basic approval for maintenance regarding technical debt has been an uphill battle.
Where do you work! :)
The nicest thing was when you nailed down the architecture and then you are mostly twiddling your thumbs while waiting for the PM guys to have the next feature idea. Maintenance and bug fixing is a breeze.
Those are the parts the LLM is bad at. The 1% of stuff that was interesting was worth looking forward to.
I’m unconvinced about architecture mattering much. I used to think that, but these days I’m leaning towards it being the product of a problem we don’t have anymore. Architecture used to be dangerous and time-consuming to redo. If an AI can do that safely for $100 or $1000 in tokens (one day), that ROI is maybe negative
I guess you’re right, though - what is the ROI of keeping the architecture clean? It certainly saves time and money later, but it’s unclear if it’s a net win.
(It’s of course also significant to note that the upfront work of figuring out the right architecture and keeping the code clean has also been correspondingly reduced as well. Both sides of the equation are affected…)
The industry never really figured out a way to reliably answer this question in the pre-LLM days either, so… I suspect it will continue to be hard to determine.
I have very mixed feelings.
- AI lets me plough through issues which previously would require debugging, troubleshooting, googling, stackoverflow
- my coding abilities are starting to atrophy
- not having to deal with low-level language quirks is nice (parameter differences of subString in C# vs Java).
- but I'm seeing the same as OP: when working on a team, the sense of ownership and community is gone. You have AI un-AI-speak for you to properly read the Jira ticket another dev created using AI. Pull requests are echo chambers between the AI used to create it, review bots, other devs using AI for PR reviews manually.
- there's little denying that we devs are feeding the very machine that will replace us
That is a problem though with AI currently. E.g. I've found a severe issue with Claude Code and prose writing when trying to make an explainer site. Instead of "plain" english it likes to use "flowery" phrases and "punchy" statements. This forces the user to spend extra time thinking about it.
And code comments are terrible. It can use 3 paragraphs to describe something when only one sentence is enough. Then further agents (or humans) have to waste their context window to figure out what the "riddle-ly" worded sentences mean.
But the things you care the most about, the AI doesn't unless you tell it to care. And even then, you have to be very careful because it acts more like a contractor with the potential to stop in the middle of a project (lack of tokens) and leave it in a very bad state until it's complete.
Or worse, it forgets what you want and does it's own thing for a while until you remind it.
The looming thing today is qualitatively different, because it's not formalizing something to reliably mass-produce.
The big money valley bros figured it out just before they embraced regressive conservatism for their politics. There’s a reason we went from massages and pet daycare to rentier agriculture executive style management of the peons.
And that reason is "service businesses are too expensive in California because of high housing costs".
Obviously this is not true for every SWE. A lot of people feel liberated now that the burden of actually writing the code is being replaced by molding it through the interaction with AI. I see people that struggle because they've identified with the writing code part a lot and that's gone now that the machine is doing it, I bet that's hard.
AI is great enabler for stuff that you don't want to be good at, it cut a ton of chore type jobs out of my workflow, but for stuff you do want to get better and care for, it's both joy killer, and if you're not careful you will only learn how to prompt, not how the stuff works.
I love code as a craft. I love spending time designing the perfect abstraction. I love seeing it come together in the implementation stage. I love the troubleshooting process and the understanding that comes when you finally engage with the root cause of the issue. I love all of this so much that I ship little, all my time is spent in process.
I also love code as a tool. I love working tirelessly towards describing a goal and then watching it come to completion. I love knowing my projects consume terabytes of data and spit out useful results. I love seeing my assumptions on the outcome proved right or wrong.
Now I can do the latter without the former, and it leaves me deeply conflicted. A large part of me is more excited than I've ever been. Another part of me is quite sad, and this is unfortunately the part where my ego lives. For now, I keep it in check by asking the LLM to describe the pros/cons of alternative architectures. That way I stay learning about the products I create.
I've done a lot of functional coding in the past (in academia, in Haskell as a hobby, and Scala for a unicorn where I was an early hire.) This probably affects how I think. But I find it quite natural to organize my code this way. After all the fundamental data structures and the hierarchies relating them are what often determine core properties of your programs such as being CPU, Memory, or IO bound.
I also work as a pretty senior engineer and my tasks before agents were usually about leading a team to implement something. I would often do the v0 implementation to make sure we met customer requirements and then take a step back and help the team scale the service up. These days the team is a lot smaller and maybe just another engineer; the agents make up the rest.
So it all wishes and prayers? no wonder software quality is down.
I ultimately think learning how other people structure programs is an enriching task and not one that should cause you to dig in and discredit differently-shaped approaches.
Learning from others is great. But there should be actually something to learn. When someone says
so long as you're not writing the code or specifying exactly how it should work.
A software is the recipe for some process. The code is the formal version where everything is exhaustively specified. Not thinking about the code means you’ve stopped thinking about the software somewhere and no longer can say if it’s ok or not.
I’ve been coding over four decades and it is clear to me that most programmers don’t know how the stuff works, and not just script kiddies but some PhDs I worked with.
For those who want to know how it works, LLMs are massive enablers. I’ve written so many debugging and inspection tools with an LLM; had a model explain the assembly language produced; had rubber duck conversations with it. I’m having such fun tbh.
You're saying that I must not be good in writing tests, irregardless on how I feel about test writing.
I happen to love the process of building software by coordinating agents. It’s an addictive form of flow state that I enjoy a great deal - similar to the flow state I get playing real time strategy games like Age of Empires. It’s a different kind of flow state than writing code, for sure.
But to say that the only people who enjoy building software in this new way are those who hated coding, technology, or problem solving is simply untrue.
I realized I most enjoy higher level problem solving but don’t really enjoy implementation. Like I enjoy algorithms, data structures, memory allocation strategies, SIMD optimization, etc. But writing the code, and debugging it is tedious and takes me out of flow.
And if I’m really interested in the project itself then the algorithms take a back seat and I enjoy the even higher level problem solving of building a cool project.
I don’t think it makes you lesser to enjoy these things more than literally banging out code, and hunting down bugs. Perhaps I’ve written enough code that I don’t find the same joy in that. Who knows.
I love writing code, but I don't love every single part of writing code. In particular, I really disliked getting stuck on some irrelevant detail, some mysterious error that had nothing to do with my code (dep issues, OS quirks, silent changes, auth failures, researching alternatives), and spinning my tires, unable to work on what I wanted to accomplish. Often I never made it through the far side; I've got a project directory littered with half-finished projects.
Having an AI to figure all that junk out for me so I can focus on the thing I want to build is fantastic. I worry about losing the grip I still have, but for now I've lost some things I loved (a lot of actual code-writing, solving nice straightforward problems), but gained what might as well be superpowers to help with some real shortcomings I had. It's been a net-positive trade for me so far.
And none of my personal projects have made me rich yet...
If someone loves building technology products that doesn't mean they were "only in it for the money". Beyond just being disparaging for no reason, you're missing the path out of the depression-pit here.
If tech paid less, say minimum wage, they would all leave and none of them would ever tinker around anything in their free time. I’m not depressed, I’m just disappointed by how many folks are like that in this industry I loved.
You can feel free to cope by inventing imaginary bad guys. But the reality is that there are a wide variety of reasons people coded by hand. And people that are different than you aren’t somehow inferior or bad.
You're certainly welcome to miss what coding was, but being rude to engineers who are trying to build businesses is just as ridiculous and entitled as this article is.
It's funny how people produce tons of code they can't even read and somehow feel like they "solved the problem". Do you feel like you built the shed when you hire carpenter and ask him to do it..?
And why school district need you when any highschooler can produce similar code..?
I don’t need to see why you feel the desperate need to larp as a technologist when you’re happy with letting an agent solve all the technical problems.
You only think it is rude because you have tied your identity to being a technologist when you are not, I’m just saying it like it is.
This is a false comparison. Agents cannot solve all the technical problems, so the entire statement falls apart.
Out of curiosity, when you take a bite of a burger do you call it a product?
When you watch a movie do you call it content?
And given that fast food is the pinnacle of sustenance as a product and Hollywood has been content-brained since 1977, sure!
In other words, people who just wanted to become managers.
At least I’m honest, I never cared for what my company does, I only liked the interesting technical problems it let me solve, and the high salarye it gave me, now I can only thank it for its high salary.
The stuff I’m building is things I could only dream of before. With LLMs, I can build things in a year that would have taken me a lifetime before.
I enjoy what I do but I confess to being “in technology” because I need to earn a living not because it’s some sort of passion project. I think it’s pretty horrible to wish us all out of job so you can feel superior again.
Regardless, I wish you all lose your jobs to AI and be forced to leave technology and earn money elsewhere. I think it’s been a disaster to the quality of the technology surrounding us and the community that creates technology to have so many opportunists just in it for the money.
Why do they have to hate coding, technology and problem solving, that makes no sense. They just use coding agents to write the code for them to use technology and solve problem.
Let's not forget that there is this silver lining of the tech. We can be free[1].
__
[1] Requiring a one-time payment of about $200000 + a small gas turbine
I venture to say that this number is small. Most SWEs get into it because they don't view writing code as a "burden".
The people who are "liberated" aren't the SWEs, it's the people higher up the ladder than the SWEs, who can't or don't want to be bothered with coding anymore. For them, managing a fleet of agents is easier than managing a group of SWEs.
I don't think it's fair to say that all SWEs swing one way or the other. I absolutely love the fact that I will never have to waste time debugging cargo culted, trendy libraries ever again while I am also sad I'll also never to get hand write elegant implementations of interesting problems ever again either (unless for fun).
I would be happy to keep writing code but I'm also happy to be liberated from the type of code I was often paid to write.
Agents help you bang out good code faster if you know how to use them. My relationship to Claude and DeepSeek is somewhere between an IDE, a coworker and a friend (I still have real friends don't worry).
I don't treat it as a world changing magic wand that fundamentally changes the way everything works. It's more like a fast forward button that lets me skip the boring part in exchange for some hand holding. No more dealing with syntax errors or finding trivial bugs.
This always read like
Writing sentences was the annoying part. The fun come from sharing ideas, optimizing for brevity, making the rhythm flow better and words more impactful, all things that are on the level of communication, and not grammar.
Code is formal writing. Expressing an idea in such formal forms means you have thought exhaustively about the evolution of the algorithm/process. If you’ve not done that, you can’t guarantee the correctness of that algorithm/process. Is it just toying around with an idea of software?
In the corporate setting I think though it won't take a lot of time to reduce the layers and number of people that are required to run a software company. Either one becomes a more general work manager or obsolete I guess.
All of your potential customers are just going to vibecode their own software. In many cases they won't even realize that the software they vibecoded even exists as it will just be an invisible tool the LLM built for itself to reproduce a similar solution to a similar prompt over and over.
Once AI flattens out any company enough, the company itself no longer needs to exist, it can trivially be replaced by the model providers. Companies exist to be collections of skilled people with a common goal. When you don't need that, you don't need companies. The agents can work for your customer instead of you and nothing changes except who gets paid (not you).
There's a ton of magical thinking out there where people seem to assume AI will progress right up to the point where it is maximally beneficial to their job without replacing them. Keep huffing that copium, folks.
I've commented previously that there will be much more political interest in AI causing unemployment when the first CEOs are replaced.
That's going to happen. It's part of the "creative destruction" of capitalism. If AIs maximize shareholder value better than some CEOs, those CEOs are toast. They either get replaced in the same organization, or a new company arises and crushes them.
Just imagine what would happen if JP Morgan did this and the model decided it needs to dump their position in AI data centers. Or maybe it's a health insurance company that fires all of the claim deniers for being unaligned with Dario's ethics.
Because you are singularly right about the issue? Have some intellectual modesty, brother.
I love writing code. But I also love solving problems. AI allows me to solve problems in different ways. And I am totally enjoying that as well. If someone got into this SWE only to code, yes, they may not love where things are today and appear to be going in the future.
I don't know if it helps, but they should know that for every one of them that experiences this grief, it brings joy to ten other people that are now able to do things (make vibe-coded programs) that used to be out of reach for them.
imagine spending years to develop mastery of something that anyone can do now? For example, sitting your ass down in college, debugging lines of code on a Saturday night while people are out partying? It's fun to do, sure, but there's still a huge opportunity cost.
It's easy to say that such people are prideful, arrogant, only cares about being better than others, but I think such a feeling sucks for 99% of the population that are not siddharthas.
The most vocal denouncers of egoism are utterly pathologic egoists themselves.
Edit: I'm at a loss for words.
There are some interesting things being built by AI, but 10x the number of joyful software engineers ain't it.
Every new technology is going to also enable bad people, just like the Internet enabled scammers, it enabled a lot more people to be more productive.
Comparative advantage ("Is this the best use of my time?") is a real thing, and the human time to make and maintain non-trivial software will always itself be real.
Speak for yourself.
Well, that is not entirely true. The British parliament made sabotaging work machinery punishable by death. So people did in fact get hurt, the people who were hurt were the Luddites, and the people hurting them was the British state executing them.
The Luddites only wanted to keep their jobs.
Brother, they're you. You just don't realize it yet.
I never cared about the social aspects of it other than releasing the odd plugin I developed.
My social life is entirely offline.
We had our work stolen and they're squeezing us with it, all for garbage, what in the world are you talking about.
Well, no, for a lot of us, there really is no other side. Slop farming is not software engineering and it never will be; scribbling some Markdown that begs agents to omit bugs is not equivalent to using high-level languages with thoughtful abstractions to directly define program behavior. I didn't want to be a manager when I worked in the industry and I certainly don't want to be one now. Moreover, it's downright offensive for a purported expert to defer their expertise to a SAAS black box.
On the bright side, we will probably (eventually) self-segregate into communities that code however they want and leave the other side the fuck alone.
What's the point of contributing to OSS if anyone can just LLM wash your work knowingly or unknowingly? You won't even get credit. Heck, you won't even know if anyone is using it!
But then what if you have a larger project than a library. Do all the leg work figuring out what works and what doesn't. What users like. Then anyone can just point an LLM at it and clone it for $50? Fun.
Those of us who refuse to use that stolen labor are loosing our jobs, fast.
I'm sure word is getting around of MBAs and etc getting access to GitLab and using AI to make python apps. Look, you don't need devs anymore?
Okay, but wouldn't a software engineer be better at that? "No, because I'm an SME, I know finance!". Right, but the LLM knows finance?
A year ago these tools were assistants, copilots, but know.. It's clear if the pace doesn't plateau or slow greatly we are headed towards replacement of massive swaths of knowledge workers.
And it risks completely destroying our society due to the economic fallout from mass unemployment..
-- Paul Krugman, 1997
Note this does not say "it will decrease if we suddenly invent new productivity technologies".
I am fine with people using my work, copying my source code, and distributing derivative work. I am not fine with AI companies stealing everyone’s work and renting it back to us.
Did they steal Wikipedia? Wikipedia is still there.
This would be true even if I signed a contract stating that I give all my labor away for free to my exploiter. You cannot sign away your rights like that, and my exploiter is not allowed to abuse my goodwill this way.
FWIW, many of us who do open source would also find that hypothetical world in which the idea of copyright is abolished rather enticing... but you pretending that is this current world whenever you want to steal our code whilst copyright still paints our interests in using and altering the code from companies that have always been in fear of our community and our licenses (Apple, Microsoft, etc.) as theft is not merely a bit shitty... it is kind of evil/cruel.
It's a bit like Bob Ross' Joy of Painting: if everyone can paint then anyone can make any painting they want to see.
the functionally same product is still being made, but there is less human involvement in the craft, and more of a human management over the manufacturing process
Probably the LLMs saved me because deep down I knew that my own power by itself wouldn't be enough to let me do what I really want, but right now the future does seem bright.
"right now the future does seem bright"
This reads like "Fuck you, I've got mine." Who cares about those wage workers anyway, right?
Anyone can do work on their own car. It's not especially hard.
But a professional working on their own car knows tricks and has tools / ways of working that a regular person doesn't have access to.
You can tell the difference in the result if you know where to look.
Sometimes it matters: the way you change some components matters - and while you're there with half the engine bay cleared, you can fix a few other things that usually fail in this specific model.
Sometimes it doesn't: oil change is hard to fuck up, but not impossible.
The the things I've made with agents aren't hard to do for anyone, but I can claim that I can pick up any of my projects 6 months from now and confidently add features without worrying about stuff breaking or the agent doing crazy crap - because I build the limits into my projects by default.
For the vast majority of software that isn't the case at all.
If not all agents are equal and require calibration, then it seems like people can specialize, so you're only competing with other specialists in that area (and not necessarily trying to out-compete).
This idea that there is 0 skill to using them is the opposite: it requires far more skill because the human has to manage the whole building all at once rather than focusing on putting up one beam at a time, one rafter at a time. Your responsibility is totally different. It has to scale far higher!
Many devs have had strong leads that do the ticket work, that figure out what to do. It's unclear what role there is for those folks now.
But none of this is going to bring back actually writing code for the love of it. That will become a hobby kind of like knitting or sewing.
Or during a sales meeting, an experienced engineer can have a quick prototype done to show the client.
Another PR and technical debt.
Being able to program means in particular, you can read those programs and understand what they do.
What happens when you don't understand them anymore?
Living in a reality where everything is a black box, doing its thing as if by magic, leaves you in a position devoid of control. The objects upholding your life are incomprehensible to you, you cannot change them or even only reliably exchange them with alternatives.
- Eric Hoffer
And even if you understand how it works, how does it affect your day-to-day function or the advancement of the field? As Apple liked to say, It Just Works (TM).
Would you be willing to drive one if the engineers at ford were all fired and an almighty agent who kinda did the almost right thing most of the time was designing the black box?
So yes, maybe we will all just exist to get sued at the end of the day.
Instead of driving you're telling a piece of nondeterministic software where you'd like to go and then watching the result as a passenger.
For me it's about losing much of the entire career, because companies will come to expect AI-level work that I can't/won't provide.
> Being able to program means in particular, you can read those programs and understand what they do.
IMO This assumes that one views programming as more than just a job/career, but not everyone does so.
> Living in a reality where everything is a black box, doing its thing as if by magic, leaves you in a position devoid of control.
For some, this is actually relieving. Sometimes we want to be more ignorant, sometimes it makes us happier.
The point isn't some people not understanding whatever stuff with limited effects, it's no people understanding anything important.
Mathematics is about describing fundamental reality, so you can change it. When corporations own it exclusively, they can do whatever they want with it.
The chilling bit is that the machines are doing just fine on the understanding programs front (up to and including decompiling them from machine code).
Now, what you're describing is really about the "culture of convenience". People prefer Expedia to human travel agents because it reduces friction (of interacting with people). Same goes for Uber, Waymo and so on. Good piece on it - https://www.cbc.ca/radio/ideas/cult-of-convenience-9.7371392
Reverse Engineering became much easier, it is plus. Finding vulnerabilities also became much easier, another plus. Models are getting better. But where is the new cool software ?
Edit: I will go even beyond OSS. M$ & Goog, who are DEEPLY Vested into AI, have access to the best possible frontier models and large groups of highly proficient engineers. Yet, Windows and Android are as laggy & buggy, as they were 4-5 years ago.
My life has been improved immeasurably (our home is filled with attachments and affordances built on-demand by Claude), and no new software was created that anyone else would see. It is entirely possible that these tools fail to create a Photoshop/Blender replacement and yet replace Photoshop/Blender.
My wife started out building in Blender[1] but these days even she uses Claude Code and mostly ideates. The nature of LLMs might not be "they can build you something better than Photoshop in a day" but that "where we're going we don't need roads".
0: https://wiki.roshangeorge.dev/w/Blog/2026-01-11/Modeling_Wit...
1: https://wiki.roshangeorge.dev/w/Blog/2026-01-15/Modeling_Wit...
Open source has had a QA problem pretty much its entire life... not in the sense that it stinks necessarily, but in that getting it at all is a challenge and was in many projects probably already the blocker to further progress.
One wonders if what is happening here is just that even in the software space, where the fluidity and flexibility is maximized as we work with nearly entirely digital constructs, that once we burn through the dry tinder of "all the stuff we know we needed but didn't have the coding time to get to" and the frontier of "work to do" advance on all fronts to "some sort of human or real-world blocker", it may just turn out that the steady state improvement in productivity after accounting for Amdahl's Law, even when code is "nearly free", is only a maybe 2x or 3x improvement after all. If this is true, goodness help the financial bubble which is assuming that something like 5x or 10x indefinitely is a lower bound, if even in the best possible domain for AI that's all we really get.
At most places I've worked, the "designing and writing code" part (basically the part that LLMs can do) was maybe 5-15% of the calendar time it took to actually get the software out into the market. The remaining 85-95% of the time was spent on squishy people-things like aligning stakeholders, getting executive buy-in, working with x-functional teams, getting priorities aligned so one team wasn't done in 30 minutes while the other team took 30 days, doing security reviews, doing privacy reviews, doing legal reviews, doing QA and verification, lining up marketing, educating salespeople, waiting for people to get back from vacation, waiting for people to get back from sick leave, waiting for people to respond to their E-mails, waiting for people to click "approved" on various internal review sites, doing final reviews with executives so they know what their company is doing. These other things take up the vast majority of the time it takes to release a feature. LLMs are only speeding up the relatively tiny 'code writing' part.
It doesn't seem like the SWE labor market has been shedding anywhere near that order of magnitude of jobs (there are about 1.7 million SWEs in the USA alone) or that companies going heavy into AI are boosting revenue significantly.
I think all things considered we're only getting a single digit percentage total productivity increase.
I wish!
The quality of commercial software has really gone downhill the past decade. Some of it may be AI, but it has likely more to do with the fact that fewer and fewer want to pay for QA.
On the endless framework treadmill, money is better spent going forward and changing things.
You point all these new software but some of them are literally only a few months to a few weeks old and complaining that they aren't up to standard yet. Even if LLMs didn't get any better than they are right now, these projects have barely even begun, and there is plenty of time and resources to build them up.
I've written a bunch of cool new software with LLMs in my spare time -- are you sure you even have a way to notice all the new cool software that exists?
This might be a problem in itself. I find myself unable to have the time to even investigate new software because even newer software with even more features comes around a week later. I bookmark so many Show HNs that I'm impressed by and only ever go back to a rare few.
I personally (no LLM) wrote some open source software a few years ago that maybe has a dozen to a hundred users and somebody used an LLM to port it to MacOS and Linux.
But even so, the entire ecosystem has been changing over this entire time. You might be using an entirely different set of LLM tools now and/or those tools themselves have gotten a lot better and easier to use.
A lot of these harnesses and mcps themselves are built with llms.
No matter what you think about it, it's impressive what the folk at Omarchy[1] have been doing with LLMs.
From what I can see, it’s just a bunch of defaults pre-installed on top of Linux. Some of the defaults they use like Proton is far more complex and impressive than their entire distro , and that was built before agents.
Then read Omarchy news.
It's here already, but not at a global scale.
You can pick any application you pay for, but only use 10% of the features of, and create a bespoke version that only has that 10% in it. It'll take you a day to get an MVP and maybe a week or two for it to be "good enough".
"ChatGPT, list the files in this directory. use ls"
vs
"ChatGPT, list the files in this directory. no utility exists but here's a hexdump of the hard drive"
the AI might muddle through with option #2, but #1 is way more efficient
I would say Photocraft is around 75% of GIMP or Affinity level and already better than both in some ways, like speed and memory usage. This is less than a month after first release and if you remember than one you know it was rough.
Of course the last 10 will be the hardest, but 90% of Photoshop for free with native PSD and lower system requirements is a win no matter how you put it.
I think the models are good at this because we can just set them loose and they can permute through their training and because a lot of things are common patterns, this can work, but is also generating massive amounts of slop that people are dumping on maintainers who are getting overwhelmed by it. I really want to know how many failed iterations a successful run cost.
One unanswered question is if this is economical. Like, if this is just gobs of money and is unsustainable, what did we just spend the last few years doing? I think these things have value, but not like this.
[ 1 ] https://pixellect.app/
>When people claim that it’s no longer sustainable to write code by hand, they’re not entirely wrong, but only in the same way that it’s no longer sustainable to sleep in a bed once someone has taken a shit in it.
The carpenter is free to exercise his craft as he see's fit, but it's going to take place mostly in his personal workshop, for personal reasons, not to develop homes or run a business in the way he did before.
We can talk for ages about the quality of the robot's work, the maintainability of the resulting home, etc. But it's clear that, at this time, the market and ecosystem are embracing the robots whole-heartedly so those of us used to the old ways are having to rapidly adjust.
The truth is no one knows how much the robot costs, and we probably won't know until its too late. Its extremely naive, though, to simply buy into the PR from the peddlers and redistribute their propaganda verbatim in the way you are doing. Although none of us know for sure, considering R&D, infrastructure capex and externalities, the robots are most definitively _not_ cheap.
OK, so don’t buy their PR, and try it yourself.
But if you do this, you will find there is a lot of evidence for the economic argument.
I tested out Google Argon 4 release by having it build a Web app to replace a binder I have for reconciling payments with reimbursable expenses.
It did it, as well as I would have and in much less time. I just set it to high and supervised it, it cost about $2. Even if you throw in a “supervision cost” to account for the time I had to spend interacting with it, it was substantially cheaper than had I done it at normal salary rates. Probably 2 orders of magnitude cheaper, even.
All of this stuff is so new, and we're supposed to hand wave away quality and maintainability?
For corporations, sales, probability and feature release cycles take priority over maintainability or code quality. AI is allowing businesses to build features faster than ever, so management is happy.
By the time maintainability issues hit, they've either pivoted, raised more capital or can throw money at the problem.
My initial reaction to reading this is not "oh great analogy", it's "please have higher standards for your quality of living"
this is why i try to avoid analogies whenever i can. we've cycled through like 3 different iterations of the analogy in 5 comments, just to land on this question.
If someone complains they take pride in understanding and designing deterministic processes, you can't reply that this is just like when robots took over manual labor.
No, because it is not manual labor, and robots hasn't taken over. Robots are computers. Robots are deterministic.
A better comparison would be an artisan architect, a blueprint designer being replaced by a special paint splatter, which when thrown against a piece of paper looks almost exactly like a blueprint of an almost completely liveable house. Great! What's not to like? It will allow anyone to design a house.
(1) We help you
(2) We undercut you
(3) You go out of business
(4) Your buddies go out of business as well
(5) Prices start rising
Supermarkets, restaurant chains, mathematicians or programmers it makes no difference.
I wondered if many a press article seen in media re security breaches and LLMs based hacks are not in part at least campaigns to legitimize regulation.
How would regulatory capture of LLM realistically look like? Do you have any ideas here?
<https://chinaselectcommittee.house.gov/media/letters/moolena...>
Let's flip this around. What do you imagine the most likely form that regulatory capture crafted by the major LLM Manufacturers would be?
When rolling this question around in your head, remember that the US Congress gave the telephone companies that blatantly and knowingly violated FISA retroactive immunity for their crimes in the 2008 FISA Amendments Act. Congress will move mountains if the majority of its members are scared enough and/or see enough profit in something.
Eh. Transhumanism is very good. Every medical intervention that improves or consensually prolongs the life of a human is transhumanist. Our natural state is incredibly godawful. [0] The best future for humanity is one where one need not work, need not get sick, need not die... where all of the awful parts of life are optional.
What the TESCREAL cult is peddling is incredibly bad. They have folks believing that not only are they ushering in the arrival of Robot Jesus, that nearly any terrible thing done now to speed up the project is justifiable because it hastens Robot Jesus's arrival. In actuality, the only thing that's hastened is the accumulation of wealth of awful humans like Altman, Amodei, Huang, and many others.
The core tenant of Rationalism [1] is great. It's just a huge shame that so many of the folks who were attracted to the philosophy ended up being so bad at learning how to do the one thing it tries to get you to do.
[0] Remember that it wasn't that long ago that humans had a 50/50 chance of dying before the age of five!
[1] Your brain is really bad at thinking carefully. You have to train yourself to think carefully despite how shitty your brain is at it. You will inevitably fail at this, but that's no reason to stop trying to get better at it.
One of these is not like the others.
> Your brain is really bad at thinking carefully. You have to train yourself to think carefully despite how shitty your brain is at it.
What is "you" in this scenario if not your brain?
Nah. All three are covered by my original statement:
The best future for humanity is one where one need not work, need not get sick, need not die... where all of the awful parts of life are optional.
> What is "you" in this scenario if not your brain?
That's an excellent question. Are we the blob of fat that hangs out in our skull, our entire body (because of how changes in "distant" parts of it affect what our brain does), the parts of the patterns in our brain that seems to correspond to our conscious mind, some other thing or things?
I am utterly unequipped to answer that question, but hope that we collectively get around to trying in earnest to answer it. Maybe we might even do that before we get around to collectively deciding that it's a catastrophically bad thing for people to go without food and shelter? [0]
[0] Given what I see here in SF and elsewhere, it looks like we currently collectively consider it to be somewhere between "Not at all a problem" and "Yeah, maybe we should get around to fixing that some day. I _guess_.".
[1] in theory.
The difference with those is that they're systems of governance... which means that they must scale if they're to be applied to large groups of people. And -well- they scale very poorly.
Example: https://m.youtube.com/watch?v=ca7ufSI1Qxg&ra=m
Most depressing part is humans are training robots to replace themselves
You can find example of this in all of the trades. Often, these are also talked about as "safety improvements" - which they sort of are. Given the same baseline intelligence, it's harder to dangerously fuck up plumbing or electrical in 2026 than it was in 1926.
The Claude Code development team has pioneered that new way of sleeping.
I think the author is trying to say that a code base once infected with vibes, becomes indecipherable to the average human being. It's still just code, but full of Claudisms.
No, not really.
Next time someone complains about software do me a favor and compare it to a nail through the foot or grabbing a hot dish and burning yourself or something else negative in the human experience. There is plenty to choose from.
We don't need 300 variations of enshitification, the folks at Pompeii and Herculaneum have that well covered.
Will be quite funny when there's a noticeable decline in stability and development—despite having AI—because it turns out you actually need talented, creative people to build the building blocks the AI spits out. LLMs can parrot but they're basically the Crazy Town [1] of software development.
And they’re not even learning from their AI agents since they’re not closely reviewing the code.
The chance of code quality increasing seems remote.
And that’s not to mention the new generation—having rarely coded enough to properly review code.
Nor to mention the AI learning from AI coded slop: mad cow disease for AI.
What do we have, no programmers, no one knowing what’s going on inside the test passing slopy black box, except a caretaker guy who’s kept their skills sharp by not using AI—with their time taken up trying to understand constantly rewritten code?
This is also the way I wrote programs when I barely knew how to program btw and I got stuff done then.
Think about doing carpentry by nailing planks together in roughly the pattern you want, rather than meticulously drawing and then cutting outlines - which I have also done. The first iteration (of that carpentry project) fell apart the moment I tried to use it and was a miserable experience because I still had to complete the task with the pieces, and didn't have the opportunity to go back to fix it. The second iteration, which still just took a couple of days to build, was much stronger and was awesome. It can be fun and get results that are obviously acceptable because they are what the AI would have output anyway.
Actually, just for curiosity, I just asked ChatGPT how to make the thing I made, and ironically it's doing a big design up front process where it wants to know all sorts of parameters first before it will tell me anything. So my "AI mode" was more direct than an actual AI! It also wanted me to buy a welding machine even though the thing can be made just fine, better even, out of wood. Maybe that's that new product advertisement feature.
That's where we're headed but I'd say the smart teams will have that 'caretaker guy' use judgment to know when and when not to apply AI. That describes one of my businesses perfectly [1].
When?
Epic is completely different, their app was awful before LLMs and the rewrite made it a lot better. It loads 10x faster, takes half the memory, every action is instant and it finally has a clear UI. If OpenAI wanted a success story to promote Astra this would be a good candidate.
My personal opinion is that all hell breaks loose due to continuity issues, time and money waste (excessive token spend to simulate competency), and a general air of malaise to the point where problems are tolerated far longer than necessary because "the bot will get to it eventually."
I started learning to write code when I was 11, in year 2000, out of pure curiosity of how tech works. I remember when a family friend opened up our very first computer, a Gateway desktop, to change out the modem that was no longer supported after an OS upgrade and the excitement and awe I felt seeing the boards and cables eager to know more. I got back online and immediately started learning how to build web pages.
I learned Flash and ActionScript in high school then got my first job as a developer building websites for local small businesses when I was 18 while in college. Most of those were largely custom Wordpress or PBBoard installs and configs. That was my 3rd job while I was majoring in aerospace, with no insight that programming could even be a career. It was just a fun part-time job for me.
A couple changes in major, a degree in computer science, several startups, and 19 years later here I am wondering whether I really want to stay in tech at all. The joy of solving problems using tech is just dead. My tagline on LinkedIn for a decade has been "Solving problems, sometimes with code" which is as true now as it ever was, but it's hard to reconcile what's happening now with why I got into this to begin with.
It was never about the money for me, otherwise I'd have worked at any of the big tech companies when I was getting hounded by recruiters for years. Instead, I've opted for the fun challenges at startups and getting to be closer to the customer, flexing my skills across disciplines and specialties. I'm mourning how the joy and spark are fading, but I'm also mourning what we are losing in the field. Things used to be about more than just the work, and you may think I'm naive for believing that, but it's very clearly just about the work and output now. It's soulless.
I myself have been using an open source lib that I contributed a fix for and inquired about it. I didn't just drop a PR either, opened an issue, was told PRs accepted, opened a PR, signed a license agreement and then the PR just sort hung out for months without response even after some gentle prodding.
Now, they don't owe me anything and they don't have to merge the code, it is their repo after all. However what did I do? use an LLM, fork the repo with my changes and now "I" maintain that private fork and add to it as needed.
As time goes on I think these smaller packages / clients will start to disappear as people/LLMs just spit out a client library or whatever or just implement the same thing over and over. Combined with the opinion of the author I figure oss libs for npm, ruby, pypi etc will start to dwindle.
The top tier LLMs would be as good at this as a lot of maintainers, right?
In some respects maybe in others less so.
So in my situation I am using a client for a system and am using a very specific part of it even though there is a lot of other complicated code in other parts of the client. A LLM is good enough to maintain the code path I am using and of course any changes are triggered / watched over by me. In this situation I am comfortable letting LLMs do most of the code. I would be less enthused if that was outside of my control and the code changes were more often and faster.
Totally fair. The source code is open. Thanks for all the fish.
There seems to be an inconsistency between those facts that would be worth discussing. I'd be interested to hear anyone's thoughts.
It matters because doing so used to be worthwhile economically. Now it is exclusively a part-time for people who have time to do it for fun.
No man is an island. I've been watching my love for programming die as I've seen so many of my peers and engineering heroes turn into vibe coders.
Community is a powerful thing.
Yes, but only if you're very wealthy.
As always, everyone else can suck it.
Casey Muratori had a take on a podcast that I liked comparing writing code vs. using an LLM to generate code to playing guitar by hand vs. playing digital music. Although it's arguably more efficient to play an mp3, people still enjoy the art of playing the guitar.
Cognitive scientists like Margaret Boden used to break creativity into three different categories: combinatory, exploratory, and transformational. It looks like a lot of our programming tasks are really combinatory and exploratory. Yes, they can be deemed novel and creative by individuals, but collectively they are standardized. It's just like finding integrals used to an art, but nowadays Mathematica can do it better to most of the integrals. Mastering special functions used to be the key to solving tricky PDEs, but nowadays few engineering students need to spend multiple semesters learning them. Quasi-RH was considered a Fields-level hard problem, yet AI could brute-force a proof with elaborate estimates. We humans have a brain that can process only a few variables at a time and it took ingenuity and years of effort to master a series of concepts before using them to solve a problem, but that could be just the cost of knowledge transfer, which AI can drive to zero in many cases.
However, the need for creativity does not go away. It just moves more to the transformational category. In addition, our craft will shift to arguably a less romantic tier: problem formulation, verification, and grounding. It will be a painful transition to many people, me included, but we have no choice but to adapt.
Note I'm not ignoring the economical impact of AI. It's very possible that lowering the cost of producing code increases the aggregate appetite for software, but the demand of software engineers can still go down. Even if total code volume explode tenfold, the labor market may not absorb displaced engineers. I wish there are solutions to this pain, but I fail to see what they are.
Assuming that it can, and it can also do a great number of things that we previously considered creative work, what’s left?
I recently played some music for a friend by an artist I really like (mitch murder), and his comment was that Suno AI could generate plenty more songs along the same lines. Now, I’m not so sure it actually can; I think the human behind the music is bringing something special to it, but this nuance is lost on my friend. Suno AI is good enough for him.
I think the situation with these ruby gems is similar. I, personally, would rather use the human crafted gems, but there are plenty of people like my friend who think there’s either no difference or no important difference between what the creative human made and what the machine can make.
With LLMs, that route largely doesn’t exist. Yes, you could write your own harness and agentic tooling. Maybe you could RLHF an existing model. But it all feels less tangible, and you can’t really change the interface to the LLM on the semantic content level in a well-defined way, like you can precisely control the semantics of a library. What you’re prompting against is not malleable in the manner that what you’re coding against is malleable. If your LLM repeatedly goes down the wrong paths, you can’t fix that once and for all, like you can fix a misbehaving library.
On top of it you can in parallel work on the library, groom its shape and then have another workflow integrating that library into bigger library and so on.
The biggest issue I always had was integration between components. Today's that's free lunch. You can extract, abstractize, re-integrate things for free.
Long before LLMs there was a concept of Big Ball of Mud. The slop existed before LLMS and all software eventually rotten. Today just as LLMs whip out faster the code rot happens faster too.
But engineers learned way into how not to make systems decay fast and (not surprisingly) the same methods work just as well with LLMs.
I find LLMs like a blender. If you are ignorant in use everything will become a tasteless grey sludge. But it's a indispensable tool in a kitchen that makes many things simple, easy and tasty.
But also it has a bolted on type system.
I dont mind burning tokens on code that is well typed.
Just like I dont mind burning tokens on TDD.
Now I can build things that would never have been possible without huge investments of time and/or money. The latest project: a hybrid human/AI tech newsroom that leverages LLM-driven advances in translation (https://asia-tech-news.ghost.io).
For my personal passions, LLMs have completely opened up areas that were extremely painful to deal with 5 or 10 years ago. Take genealogy: I used to manually transcribe old letters, and documents in foreign languages (often archaic terms) were impossible for me to work with without seeking out experts. The other night I had an AI go through a book of 200-year-old religious records and pick out references to potential ancestors. A few weeks earlier I worked on an old diary transcription (https://youtu.be/P6r1I-4T9is?si=qhaf3Wd5zZNjieEw) which I later shared with relatives to proof and enjoy. Such tasks are no longer a chore. It's fun again.
The closest comparison I can make was stumbling upon an old HTML tutorial on an MIT website 30+ years ago. All of a sudden, as a writer and creator, I was no longer tied to my employer's business model to get my words and images out to the world.
Take woodworking. For some people, the act and meditative work of building it piece by piece is the part that they strive for. For others, it’s the planning phase and the “how will all these pieces go together” phase. For even others, it’s the staining phase. Some people like every part, others like just one part and work through all that extra just to get the chance to do that next part. For some, while they may not enjoy the preface to the work they want, they appreciate it because they feel that blitzing past it cheapens the stage they love.
Everyone is different and AI is hitting everyone just a little bit differently, and that’s just in areas that directly impact their specific craft.
That's a great point. I don't have experience with woodworking, but when I see a piece of good antique furniture with precise dove-tail joints that someone made hundreds of years ago, that brings me joy.
As a writer, and as a reader, I really value human-crafted insights, like many of the comments I see every day on HN. As many people have stated elsewhere, writing forces you to organize your thoughts.
But I also recognize that there is a place for machine-generated content, whether it's the weather report on my phone or the factory-produced chair I'm sitting on now.
I don't think the story of artificial intelligence is "humans will stop creating." AIs are powerful tools, but they shouldn't be regarded as mutually exclusive with people who like to create things. Some will create in old ways. Some will create in new ways. Others will move onto new ways of creating.
Personally I enjoyed the act of programming/building. I'm struggling to put this into words, but software engineers have often been interested in solving technical problems, rather than business problems. At least I've felt this personally. I'm working on shifting that enjoyment toward the requirements gathering/design process. At the end of the day, it's still problem solving (which is the root of what I enjoy). But now it's more focused on solving business problems vs. technical problems. Which at the end of the day, is what is valuable to the business.
From the business perspective, this is great, because the end goal of software for the business is "business value". But it is a tough shift for some people to make (me included).
Edited to add: Writing code vs. designing software are both creative. Writing code has a much more concrete/objective feedback loop though.
I also particularly hate the feeling of AI being shoved down everyone's throat.
And I don't enjoy digging through the mess that AI creates, whenever I have a particular bug that it can't quite get right.
TL;dw: DHH recently held a RailsCon keynote funeral for hand coded Rails in favour of AI Rust. It's well presented and most of the attendees are probably out drinking.
You might already have _been_ the problem here.
Fixed that typo ;)
A lot of people get involved in open source because of the sense of community or because they like the idea of giving something to the world.
What's the point in contributing to opensource if your work gets violated by vibecoders and corps that disrespect your licensing and give you no credit while simultaneously clamiing your work is useless and you're going to be replaced by a LLM model?
Any ecosystem declines when there are too many cynical grifters compared to well meaning hard working people.
I really enjoy Ruby, rails, and agent-driven development. So I’m probably not the target audience. To each their own.
The pool they're talking about isn't software or CS, it's the attached treadmill that provides social kudos and credit for a job well done.
This whole thing has been a huge revelation for me in that I naively presumed at first that other 'computer people' were also in it for the passion and interest in the hobby, now I am realizing quicker and quicker that "it's just a job" for the majority of people I talk to, the rest of the people do it "for exposure" , which is a nice way of saying that they're playing the Facebook/MySpace social kudos game with their linkedins and github profiles and blog articles.
very few Terry style "just passionate about the thing" people out there, and I had never realized how rare that was.
Entirely fair. It will be very very interesting to see what happens to software when everyone is running their own vendored fork trying to maintain them with AI agents.
Will the whole thing collapse under its own complexity? Or will smarter agents eventually dig their way out of the hole dug by their predecessors?
I believe the Ruby community counts many people who resonate with this aspect of the language and of the community. And if you joined and enjoyed Ruby because you love collaborating around elegant code with/for other humans, you're surely not enjoying the current AI trend a lot.
I'm not writing Ruby daily anymore these days, but I still feel this way, and I really do not enjoy my work much these days.
Its definitely reasonable to be sad about the loss of our craft, but lashing out at others in a vague and undefined way is not it.
Open-source was always strip-mined by massive corporations. It was main source of free labour for decades.
Unfortunately there are always people willing to subsidise these corporations with their time for a badge of "honour" and few pats on the back.
Great to see more people awaken to the fact corporations thrive on unpaid work whilst making billions for their shareholders.
I am yet to see this trickle down wink
AI algorithms were developed in 1950s, and what all those vultures were waiting for was a good dataset and enough compute power, which they saw coming in advance.
> I am yet to see this trickle down
hmm...these two sentences don't agree with each other.
The idea that advertisers and malware merchants are good stewards of the browser or email or social media needs to end. Especially when said tech leaders hold deeply anti-human beliefs.
Some of the least skilled people are doing the screaming. They desperately want to establish a consensus that anyome can do what you do now and being skilled doesnt matter...
You know why? Because they have no skills, they're mediocre and the best way they can succeed in a meritocracy without skills is to get people with actual skills to give up, or become meat proxies for the machine like yourself, then maybe you'd stnad a chance.
People will downvote me here, and tell me I dont get it. No I do, you're not skilled and you dont want to have to compete on your own merits. You want me to be impressed with you because you make things with llms. Im not.
I was where it sounds like you are a month or so ago (denial), so no downvotes here. It definitely sucks; it's like losing a huge part of my identity. I'm still grieving it.
I'm trying to deal with it by separating my identity from my job. I can still write code or learn new languages in my free time (such as it is). It's mostly a hobby now. And I can choose to not use LLMs at all in my personal life. But I do have bills to pay and a family to support, so at work I'm trying to learn how to do my new job of LLM orchestrator/babysitter. It's way less fun, but the money's still decent, so I'm hoping I can swallow my pride enough to keep going.
When I signed up for AWS conference in Sydney this year, they listed out all the courses by codes in a vibe-coded React component. It didn't have any search function, so I had to manually scroll through it multiple times to pick my courses. The whole conference was AI-first.
AI is not as good as a team of developers. I don't know if it ever will be, but it definitely isn't today. AI is actually quite terrible at what it does. But it doesn't need to be better than that. Because the point of enterprise isn't to make good software.
Do you realize it won't last long at such rate?
Insane. This is the guy who was proclaimed to be "nice" by his followers. The biggest LLM sellouts in open source are Ruby and Scala.
Ironically Oracle with Java is now being the good guy.
I'm pretty sure someone will soon call Linus a sell-out. The hate / denial / radicalisation is just sad at this point... Do, or don't, but at least don't hate the others. Be and let be. No need to call other people names just because you don't like their sand castle. There's plenty of sand on the beach.
That's not quite practiced by those who preach "the genie is out of the bottle, and we invested too much to allow him go back, so adapt or die".
Reminder: oracle is also an acronym, and you might want to take a gander at what its subject is up to.
There is no future in which oracle is the good guy no matter the axis. Least of all the subject of AI.
He started being ambivalent about it maybe two years ago, and reached metastasis in the latest Rails World fiasco, with dhh telling him to his face that his life's work (Matz's terms) is irrelevant now.
I'm not sure why he isn't rich enough to walk away.
Is this really the way you look at open-source libraries that are given to you for free, i.e. as a gift? I hope you look differently at gifts people give you in real life.
> thrownaway561
Interesting.
https://en.wikipedia.org/wiki/David_Heinemeier_Hansson#:~:te...
https://en.wikipedia.org/wiki/David_Heinemeier_Hansson#:~:te...
And even with actual illegal immigrants there's a different between saying a country can deport them, should deport them or has to deport them based on some bizarre fascist ethics.
I have no problem with people who don't like immigration and want to send them back (I just really don't agree with their opinion), but DHH is closer to the last group. He's saying that each country has a moral duty to deport immigrants and every individual citizen has to support that, and if they don't that means they're evil and want to destroy their own people.
Using violent language is just a cherry on top.
I don't know anything about this person, but he seems to believe they're immigrants, because deportation otherwise makes no sense. I've never heard the word used to refer to expelling natives. That would be exile or banishment.
>I have no problem with people who don't like immigration and want to send them back (I just really don't agree with their opinion), but DHH is closer to the last group. He's saying that each country has a moral duty to deport immigrants and every individual citizen has to support that, and if they don't that means they're evil and want to destroy their own people.
I don't really understand why you would care about the distinction. Do you have to interact with this person on a daily basis?
The list of things you can get canceled for is growing endlessly.
There is sharply reduced benefit to the maintainer, however as they are now buried in slop bug reports and PRs, and a bunch of people with no context and a Claude subscription think they can do it better.
I expect a lot of open-source maintainers will walk away from that particular bargain in the near future
> You want changes? Go ask your agent or something. I’m told that’s what you’re supposed to do now. I’m out.
This is way too spiteful to use as the closing message. Yes, we will ask AI agents, and chances are the code they'll write will be better.