I was chatting with a friend who commits on a very well known open source project (~80K stars). He is disillusioned not because the robots aren't good at what they do -- they totally are. The sad thing is that he enjoyed writing code and working with the community on improvements because it was a very human, collaborative thing to do. So, while the robots are great at pushing the project forward even faster and doing fantastic things for security and performance, the joy is gone.
I can totally identify with that feeling. It must be how professional mathematicians are feeling as well.
Everyone in SWE (who got into it for the love of writing code) is experiencing this. 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.
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.
Like many, I became a SWE because I loved writing code. Ruby, in particular (whatever one thinks of its other strengths and weaknesses) can be very expressive and beautiful.
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.
This is super great. God, I wish all leaders had half this much empathy! Good on you.
I am one of those weirdos that found joy in business software. The clear specifications, reasonable timelines, and enough technical freedom, let you focus on clean, secure, resilient architectures.
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.
> The clear specifications, reasonable timelines, and enough technical freedom, let you focus on clean, secure, resilient architectures.
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.
Wow. That was hardly ever the case in my experience in ~30 years in the industry. I enjoy(ed) what you’re describing too but… how common was this for you?
Different person but I’ve been a SWE about ten years now - my favorite work has been improving the CRUD stuff people have left behind. Like the time I took a proprietary document they were using to document their REST apis and wrote a converter to turn those into OpenAI spec compliant yaml files, taking a single tenant application and making it a secure multienant application so instead of 80 different app services we had two, coalescing all their data sources into one normalized DB. There’s a lot of work I’ve done I’m really proud of. I’ve always found myself being the “in the muck” guy fixing up and bettering legacy code, I’ve always felt that was really rewarding and interesting.
Thats badass! I like doing that kind of work, but I got very jaded about it over the years, because management generally acted as if the engineers took a shit on the conference table in the middle of a Zoom call with their most important customer whenever anybody attempted to address technical debt, as opposed to delivering more more more features. Unless it was billable agency work, in which case they’d hate it even more.
With massive amounts of effort and strong communication it was usually possible to get some tech debt time, but management acted like they were doing YOU a huge favor. It was kind of a career killer, generally.
Alternatively it was possible to get this kind of work done on my own time… but that’s a marriage killer instead of a career killer.
The fun thing is: I don't think this will change that much with AI in the mix ;)
Shitty devs without AI are slowly building BS that wants to be cleaned up.
Shitty devs with AI are much more fasterly building BS that wants to be cleaned up.
I have a similar experience as yours, just add another 10-ish years to the timeline. Literally my first job out of uni, I got to work on a legacy code base that was another 15 years older than that which had a ton of reports all built differently over time with slight variations in both code and end user output that really all wanted to be "the same" both on the code and end user side. So I slowly but surely, one user/PM requested change at a time would build out a common layer of abstractions and shared code that both made the code shared and easier to maintain (great for me!) and consistent for users (great for the PM/users).
Heck, I built a rudimentary TransactSQL parser that'd output stored procedure call graphs to graphviz and then I printed it and hung it on the wall to get through all this and having an LLM to just make a mermaid out of this in 5-ish minutes would definitively have been faster than I was ;)
With an LLM all of that would just have been faster to do but still worthwhile and useful on both levels. I.e. both code abstraction/reuse and PM/user level. Similarly there was a report that one of the devs had to actually run manually and it took him a full week, so users that wanted the report would only request it once a year-ish. When I was asked to learn how to run it from him I was appalled and just took that week to put it into a script that we ran and emailed the report out once a month automatically. They ran that report manually like that for like 10 years prior to that :mindblown:
None of this is an "AI thing". It's a good dev vs. not so good dev thing. And none of it really has to do with coding as in "writing source code" as much as "solving problems for humans" somehow and the "language" you do that in changes over time. If you're the guy that's tied up with the particular language: yeah LLMs are gonna eat your lunch. If you're the problem solver type that uses whatever tools they have available, you're still in demand.
Sure. If you get pleasure in fixing really badly broken shite, there’s tons of opportunity for that, but no money in it. Only money in new features, which is why the codebase is what it is. I have, on rare occasion, been the fixit guy called in temporarily to fix another teams work or get it out the door: that was fun. Short lived , but fun.
It's a shame, because there is money in it. It's just (like most sensible preventative maintenance) that the financial benefits are typically delayed. And when it comes to software, they're notoriously impossible to quantify.
Generally the only way to get it done in a corpo environment (and receive any kind of thanks or glory, as opposed to just torpedoing your own career) is to let things get so bad that production stability suffers and/or the team's ability to ship features is severely impacted.
...I realize I'm basically just echoing what you said, except more verbose. I'm venting!
About 80% of the time. It wasn't all about setting a new foundation, even making the existing one solid enough so you can have time to refactor the cruft was enjoyable. I had more trouble in my youth, as I didn't understand why the business didn't change to accommodate my 'masterpieces', and I'd like to think I started enjoying it more as I became wiser.
Disclaimer: I had to take an opportunity at management in cybersecurity, and haven't coded in about five years. It helps a lot when talking to developers about risk mitigation, though.
One of the fun things to do would be to use the 8 hrs of work to spend about 4-5 hrs working on deliverables, which was usually sufficient to exceed expectations, using 2-3 of the remaining hours (I’d keep 1-2 hrs for myself), to make significant architectural improvements.
I’d bring in an improvement every 6 months or so and suddenly the app would get significantly better and I’d get credit for it.
> The clear specifications, reasonable timelines, and enough technical freedom, let you focus on clean, secure, resilient architectures.
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.
Sorry for the delay. Mostly midrange (AS/400) consulting at manufacturing plants, then full time jobs at a heavy machine dealership, and a bank. Changed to PC development after about 15 years of that, then management, and it hasn't been the same after that, ha ha.
I think the technical freedom plays a big part, not because you're chasing the shiniest, but because you're choosing the tools that will hopefully outlast you. One of my former employers was 'modernizing' their IBM i (former AS/400) systems by porting them to Java and Windows Servers, and they were having all kinds of problems. Couldn't even get the EBCDIC to Unicode conversion quite right. I was lucky not to be there anymore.
I smile whenever I see EBCDIC, probably because I never had to deal with it.
But I'm sorta old and some of my school's texts were outdated so they had programming examples and explanatory text that treated EBCDIC on equal footing with ASCII.
So unlike anybody who dealt with EBCDIC conversions in the wild, seeing that acronym actually takes me back to happier days!
> not 99% tweaking reports, trying to implement arcane and contradictory business logic, or fighting with shitty and frustrating externalities.
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
In a big production codebase with lots of active customers and data, architecture changes still feel pretty difficult even with LLMs accelerating them by an order of magnitude — migrating and backfilling data and all of that. That’s more complex than pointing Claude at a repo and saying “rewrite this in Rust.”
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.
in my experience it takes more than a day to sufficiently migrate bad architecture to a good one. like, many more days than it took to vibe the original implementation. having the prototype is good to understand what you want, but then going back and rebuilding it properly is a longer task. which is fine.
I think this is not "parallel universe". Selection bias: people who love to code often choose non trivial frontier work. They tend to grirf about agentic workflow the most. Myself included.
I got into SWE because I like solving problems. I like solving problems using code. I also like solving problems using AI.
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
There's a lot of code I never wanted to be an expert on. I might care about an algorithm. I might care about the UX. There are other things I might let an AI tackle because I don't care about them.
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.
We were always replacing ourselves: Every "reusable" function or class or library you ever wrote represents lines of code won't get paid to write again. Every compiler takes jobs away from assembly coders.
The looming thing today is qualitatively different, because it's not formalizing something to reliably mass-produce.
> We were always replacing ourselves: Every "reusable" function or class or library you ever wrote represents lines of code won't get paid to write again
In theory, but not in practice. Instead, somehow we ended up writing essentially the same app over and and over, usually without reusing libs. You know because this time we'll use vue, or svelte, or phoenix, or htmx, or go, because this time we actually know what we're doing. so gotta start from the ground up and figure out how to store a row in a db as if we haven't done the same thing a million times before.
And that all under one roof. Now multiple that by thousands of roofs.
At least the devs were having fun and making good money writing all that code.
It’s a point that isn’t made often. AI is mostly saying the quiet part out loud.
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.
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.
Yes. The people that didn't enjoy coding say that.
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 enjoy coding deeply, and my feelings are not that simple.
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.
As I like to tell others: before my job was 70% architecture & patterns and 30% coding & algorithms. Now my job is 100% architecture & patterns. While I still spend the majority of my time on architecture & patterns, and while I enjoy this aspect more than coding and algorithms, I am still working a "different job" than I was when my hands steered the code completely.
People say stuff like this all the time, but what does it actually mean? Like, isn’t the details of the code, at least at some level, part of the “architecture”? I’m so curious about what people who say this are actually making where they just sit around all day and dream up new lego structures without ever having to look at any of the bricks.
The bricks are pretty standardized? I mean, you're not reinventing singly-linked and doubly-linked lists from first principles no? When I program I think in terms of data structures, data hierarchy, and algorithms. I want to use a queue here, a lock here, a double linked list there, an array list over here, maybe a hash dictionary with linked-list chaining for collisions. It's only a few times in my career that I built a bespoke data structure from the ground up and they were all for very hot loops and one case involved working on a lock-free algorithm.
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.
I may have it wrong, but I think 'architecture' can include relatively low-level details: you could even design a program down to the level of what each function does, and still only be doing 'architecture', not coding, so long as you're not writing the code or specifying exactly how it should work.
> you could even design a program down to the level of what each function does, and still only be doing 'architecture', not coding, so long as you're not writing the code or specifying exactly how it should work.
So it all wishes and prayers? no wonder software quality is down.
Software is built on wishes and prayers. Programmers don't ship pixels to be installed in monitors, nor do we (mostly) work with compiler internals. We write a detailed set of instructions and hope that a bunch of systems we don't necessarily comprehend follow them well enough to get the intended results. It works because some folks got together and wrote down their wishes and prayers so good that others were able to understand and follow them later.
We can be curious about how others approach problems or we can try to find the faults. I feel like this comment is a great example of jumping to the negativity rather than trying to understand a different perspective.
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.
> I ultimately think learning how other people structure programs is an enriching task
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.
It means they’re not creating software. They’re just playing with randomly generating something.
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.
What's the point of leaving a comment if you're not going to change your mind and they're not going to change theirs? Is it just to advocate for your position?
As you're being very vague, I write more about my opinion, which is based on a guess of what your approach. Basically the Q of P imply Q where P is the guess (and P is the thing you're being vague). I'd like if you can prove that either P is not valid or that there's a different Q that exists.
Axiom: A software is the recipe for some process. The code is the formal version where everything is exhaustively specified.
P: Not thinking about the code
Q: means you’ve stopped thinking about the software somewhere and no longer can say if it’s ok or not.
To further go with my point, it's like explaining how to make a cake and you said that instead of a recipe, it's fine to provide just the name of the ingredients and a general instruction like: mix everything together and put it in the oven.
When we write code we sit atop a tower of abstractions. You're trusting your language's stdlib, its compilers, the syscalls offered by your kernel, the virtual memory implementation of the kernel. Your coworkers wrote code that you only interact with as a library or over an API of some sort. Let's not even think about the lossy abstraction of taking the ultimate bytecode generated and actually mapping it to the actual CPU pipeline that's running the code.
Modern computing is an exercise in trusting abstractions and then zooming into the abstractions and figuring out what went wrong when something breaks. Just because I wrote a web route doesn't mean I've audited the code behind the web framework, the event loop, the HTTP implementation, the TCP implementation and so forth. By your argument then I've already stopped thinking about my software and can't tell you if it's okay or not.
I'm sympathetic to both the perspective of sadness from the loss of hand coding and the perspective of joy from working at a higher level of abstraction, despite its costs in understanding. It actually seems very simple to me to hold both perspectives within me.
The thing you’re forgot to mention is that the trust in the layers below your code is not blind. Sure you treat it as axiomatic foundation. But when something breaks and you’re sure it’s not about your code, then you can investigate the actual part that is wrong in that foundation.
Once that’s done, there’s a matter of accountability where you can report the issue to the people responsible for that. Then because they know their own code they can repeat what you did and try to find where things went wrong.
That chain of accountability is the main issue there. We still have tower of abstraction, but no one knows about their layers. Accountability is lost and with it the trust is gone.
I never care about code, whether you wrote the majority of it or it’s just a glue that hold several library. What I care about is that the semantics of that module that I interact or build upon can be trusted, and when something wrong is an accident and not a daily event.
Even when you’re writing an application and not a library or a platform, you’re not free to write whatever. Your users depends on you, and themselves may have clients that depends on them. Being cavalier about what you write is just recklessness and in any other (real in some countries) type of engineering, it would be labelled as malpractice.
> That chain of accountability is the main issue there. We still have tower of abstraction, but no one knows about their layers. Accountability is lost and with it the trust is gone.
Right this is the main difference between our viewpoints. I don't work on completely open systems. A lot of my life I've worked on closed platforms, cloud runtimes, a mix of work that's been in my control and out of my control. I've filed my share of AWS tickets but ultimately most of them ended up not fixed by AWS. I've made peace with not having accountability for layers of the stack and am comfortable assuming that there are bugs on the foundation that I have to work around.
> Being cavalier about what you write is just recklessness and in any other (real in some countries) type of engineering, it would be labelled as malpractice.
I just design the fun part and let LLM deal with the chores.
For example recently I did tiny little hobby project, taking metrics of ESPHome/homeassistant MQTT and pushing it into prometheus. Coded the main parts, told LLM to do the boring part of setting up test suite, guided it thru making tests, it worked fine, tested it, then told LLM to do another boring task, taking all the datatypes it support and unit conversions (which is hundreds).
I got to do all the fun design stuff, but had code review assistant that also did all the boring tasks.
It's like making animation but you only have to make keyframes and initial coloring, with AI filling the blanks, just cutting out the tedium
> and if you're not careful you will only learn how to prompt, not how the stuff works.
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.
That's a very simplistic view. I'm very good at writing tests and I'm happy to offload writing tests to LLMs, as long as they follow the structure I've provided.
You're saying that I must not be good in writing tests, irregardless on how I feel about test writing.
People who hated coding, technology, problem solving and only were in it for the money love managing a bunch of agents. Unfortunately, it seems like a lot of them were in technology. I hope agents quickly become good enough to completely commoditize any supposed function they perform, and forces them to leave this industry.
This is an extremely “all or nothing” take. I love computers and started programming in elementary school. I am having more fun now building software than I can remember.
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’m glad you mentioned flow state because that’s exactly the high I’m trying to chase with AI and why I like it despite programming since elementary school as well.
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.
You can love technology and still have thought that coding was a required path to get to technology and not the ends unto itself.
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.
It’s not the act of coding, you have to really hate problem solving, to love managing agents. Such people are only in it for the money, it may sound disparaging but it’s just true.
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.
There are infinite problems and agents help us solve problems faster. Sure they're not coding puzzles, they're actual problems actual human beings have, but they're still problems. Id way rather help a school district track its roster properly than argue with eslint. I'd rather get a rocket to the moon than triage garbage collection quirks too.
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..?
If agent did all the technical work of sending a rocket to the moon, and all that was left for me was deciding whether a rocket to the moon should be built or one to mars, which is better for our “customers” and I actually enjoyed that then I would be honest and say I was never into technology, which is completely fine, majority of humanity isn’t into it.
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.
> 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.
This is a false comparison. Agents cannot solve all the technical problems, so the entire statement falls apart.
What you and I are saying is synonymous. I’m simply reusing language that the post I’m responding to used. And also de-glamorizing/de-romanticizing things a bit, sure.
And given that fast food is the pinnacle of sustenance as a product and Hollywood has been content-brained since 1977, sure!
I’ve coded for 30 years. But I’ve always been motivated by outcome, imagination, intuition, & speculation. Coding was always a means to an end not the end itself. This is the funnest I’ve ever had creating software.
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.
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...
I got into software later in life because I wanted to be able to send my kids to college, have healthcare and not destroy my body by 50. Up until my 40s I was a landscaper who had been an amateur programmer and technologist since my teens. I worked really, really hard to get where I am. I went to a bootcamp 5 nights a week for a year after breaking my back all day at my “real job”, while also taking care of a wife and two kids.
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.
It says a lot that you felt the need to recite an emotional sob story to try and attack my comment.
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.
Your response here says a lot about who you are as a person. Maybe it's time to step away from the keyboard and do some self-reflection and growing up.
I wasn't "attacking your comment", just trying to get you to relate to me as a human being. But I should've known that would be a futile exercise.
I'm good at my job, I like what I do and I work hard six years into my career. I just don't go home at night and do it as a hobby too. I know there are people like you would like to see me kicked out of the clubhouse for having needs and pursuing a career in one of the few middle class jobs left that pays a living wage. Sorry I'm not cool enough for you.
Why did you expand it from what he said? I love technology, I enjoy being able to solve problems with technology and coding. I've always felt neutral about the actual act of coding, it was always just a means to the end and now I love that I can do more problem solving in less time and with less toil (from my view). But I guess I was just in it for the money.
> People who hated coding, technology, problem solving and only were in it for the money love managing a bunch of agents.
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.
FWIW I hate huge chunks of the industry + its inhabitants and am somewhat happy that my GPU can replace a significant amount of interactions with it and them.
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
> 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 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.
Most of the code I was paid to write is a burden. I've spent far more hours writing generic CRUD apps, debugging overly complex javascript tooling (almost always out of date) and other tedious grunt work.
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.
For me typing out code was always the annoying part. The fun comes from fixing problems, optimizing for speed, making the UX flow better and visuals look nice, all things that are on the level of software not code.
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.
> For me typing out code was always the annoying part. The fun comes from fixing problems, optimizing for speed, making the UX flow better and visuals look nice, all things that are on the level of software not code.
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.
Then you start wondering about the actual level of education of the person and their claims about writing a nice novel.
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?
sure, it works like that if you're still the one doing the thinking and figuring out how to do things, and just using agents for the boring templating and to translate your pseudocode into syntax-free real code, without you having to look up the arguments for every call that you haven't used recently, etc.
but most SWEs are being pushed to delegate the _thinking_ to agents.
so instead of "here's a solution to load this page faster that I came up with: X. Implement this in JS", which is along the lines of what you describe, most people are doing "this page needs to load faster, what's the best way to do that?" claude: "X". "OK, implement X." What fun is that? What's the point of developing skills and acquiring knowledge if you can't use them anymore. that's the loss.
If all you care about is having pages that load fast, then yeah, agents are great. If you care about being the architect who figured out how to load pages faster, then agents take the wind out of your sails.
Your example is great because I literally just did this:
"you need to improve performance of the graph especially rollover and highlighting it's VERY Slow there are too many updates"
In the past I would have cracked open a profiler, and spent a couple days probably tuning this thing. I probably would have gotten a big dopamine hit and felt really accomplished when I finally got buttery smooth performance.
But I'm happy to delegate this task to the agent, because this is one step in 1000 steps I have to accomplish, and no I can't just tell the agent to accomplish them all from the start because I don't know where it's going. I'm very very happy I don't have to waste a day making the graph load faster, because I can get back to my real work, which is not making graphs load faster.
At the end of the day, solving that problem will make someone's use of my project a little better, but it's not adding value to anyone's life -- that's what the project is (supposed to be) for. In the past, it was those little problems and 1000s more like them that would derail a large project. Today, they are mere bumps in the road, so the large project has a better chance to come to fruition.
in this context, those are really 2 different things. what people are lamenting is the fact that "I built a career on building these skills because it's something I care about doing with my life, and now I'm being relegated to QA at best, or replaced at worst". That's completely different than, "I can now get my personal project out faster, or maybe create 10 personal projects, and can have the agent take care of the parts of the project I don't really care about doing (like improving graph performance) and focus instead on the overall design of the project, more features, etc.".
It's like going through med school and residency to become a doctor, and instead of practicing medicine, you're feeding prompts into an AI and passing on the responses. Sure, on you might appreciate not having to field medical questions from all your friends and relatives ("personal project"), but that's not your career.
My day job, but I am not a software developer (I'm a teacher), although I was in the past.
I can appreciate that perspective, but at the same time: it's a job, the whole arrangement is you sign away your autonomy, you do what you're told, and people were paid more in one year than I've made in 5 for this. Now the balance has been disrupted, and that's a shame, but the gravy train can't go forever. Moreover, this particular job, software developer, has changed fundamentally again and again over the years, and those changes definitely disrupted people's careers, so although the blast radius is significant, this too shall pass.
And we're not exactly talking about doctor-level of training here, it was only about a dozen years ago that it was very fashionable to land a 6-figure salary at a top Silicon Valley tech firm after having only attended a bootcamp, for a job that was characterized as being a "code monkey". The whole thesis of the "Theil Fellowship" was that the job was doable without any formal education at all. At my institution, the CS major was the most popular because it was the shortest path to the largest salary. Now that equation has changed, no one wants to be a CS major, enrollment is through the floor.
> you're feeding prompts into an AI and passing on the responses.
Okay but is the job writing code or building software? A lot of people seem to think the job was writing code, and that's what they liked about it, but it was always the case that if you like your job it's not going to pay well. That software engineers were paid so well to do something they love is an aberration, not something to expect forever.
As a teacher, I definitely understand this. I love my job, doesn't pay well, and it's also changed fundamentally as well in the last year. So I empathize -- I used to grade student papers which was rewarding; then all of a sudden I was grading LLM output, which is not. A lot of my colleagues have complained endlessly about this, have made it their job to become the LLM "police", or have quit the profession entirely because like you said it's not how they started in this field. It's bleak.
But you can also adapt to the times. We just completed a module where students use AI to create games in Lua, and they made the best games I've seen in years, each one different and unique and fun. Before the class was "we'll spend 4 weeks and hopefully we'll get to animations if we're lucky". Now the final projects are all animated with interesting gameplay mechanics, graphics, animations, and they're all different despite them each using AI. I like this situation better.
Totally, I can't quite imagine an AI automating my use of AI tough, how it would infer what prompt I would use at any given time, the problem space is infinite. Am I adding another *arr app to my stack, are we building a SPA, a home automation plugin, whatever. These are things that I want to do and AI helps with that.
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.
It is remarkable to me how few people seem to understand it is a much smaller step for AI-generated software to kill the need for companies of idea men than it was for them to kill the need for people writing code.
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.
> 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.
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.
I don't think that's ever going to happen when models still hallucinate and prompt-refuse, especially since one of the two SOTA model makers insists on aligning their models toward a progressive ideology.
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.
> 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.
Because you are singularly right about the issue? Have some intellectual modesty, brother.
I think it's fair to expect tha AI capabilities will more likely end up far over or under that magic threshold than close to it. Given the way that technology develops, it's probably going to be over.
Its nice to know others feel this way. I do feel some kind of despair when I think about the kind of work I used to do even just a year ago compared to now. A coworker of mine took FMLA recently to figure out what they want to do next because they no longer feel even the slightest fulfillment from swe.
Interesting take and I certainly agree that lots of people think that way.
In my case I had prior experience with tech support and I saw lots of people struggling to use computers and it's interfaces.
So my joy comes equally from coding or seeing people relieved that their problems were solved.
With that mindset that I'm able to fly with Claude or codex.
And even doing my comp sci more than a decade ago many professors told that I should worry more about the outcomes and what problems I'm solving, instead of building this life game around coding only. So I'm grateful for them.
Maybe that can help other people getting out of this dread feeling whenever they are using A?
My frustrations changed, nonetheless, with AI. When it forgets cases that the LLM session and I forgot to cover, that for me is pretty obvious if I had or gave more time to evaluate and review things.
This week I realized that running a separate session and asking dumb questions about the plans and recent pull requests helps grounding again.
I have to agree, but I think some people find the joy in finding their solutions in code rather than from code…meaning that for some it is the puzzle solving that is their interest. That can be rather hard to find an analog for, especially if coding is the skill they are good at.
The lack of AI mental model between conversations or even within the conversation is a definite frustration. “No, I didn’t just decide to entirely change everything I’ve planned based on one ambiguous word…”
Your use of a separate session for q&a is actually something I find interesting in the new Granite Switch architecture’s KV Cache Normalization, where the kv cache is essentially frozen while the model performs certain other tasks in order to prevent the mental ‘conversation’ from getting polluted. However, I’m also assuming what you mean by “helps grounding” is that it helps keep the agent grounded in the context by not introducing a bunch of external information, which is certainly beneficial.
The other side is management, instead of being team leader for an offshoring team, which at least were human, those that manage to keep their jobs will be a robot manager.
I certainly did not come to computing to be a manager, in fact I already had that role and freely downgraded myself.
Of course this does not undo any grief, dismiss it, or better their situation:
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.
That just makes it feel like I wasted a lot of time. I spent years eschewing social gatherings, foregoing other opportunities, and depriving myself of things I enjoy, to cultivate a set of skills. Now I feel like they are rapidly becoming largely irrelevant (vuln hunting, kind of on the leading edge of ai displacement, not sure bug bounties as we currently know them will exist in five years time). What's more I don't feel like I'm positioned to remain employed in the field. I didn't expect my status or stability to be thrown into jeopardy so rapidly. I wish I were as selfless as you. I'm not sure many people are, maybe the world would be a kinder place if they were
that probably makes it feel even worse for most people
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.
All this supposed fun, is always something fairly simple that would take maybe a hundred or thousand hours of practice to build, it was never out of reach of them, they just felt they didn’t have the time or motivation to build that skill. (even a 1000 hour project can be done in 6 months outside of work time) Now they don’t need it. If there is any fun in this, it is an extremely shallow, garden variety sort of fun that comes from consuming content. They will never experience the deep fun that comes from being immersed creatively in a project, which I would trade anytime for something shallow.
We aren't suddenly seeing 10x the software output that these 10 joyful people are producing. I think it's more likely that 10x people are suddenly trying to figure out a way to low-effort scam money with AI. And a few of those people have teamed up with VC's to make AI products to scrape the bottom of the pond for desperate suckers looking to scam money. There's 10x the headaches for existing projects that suddenly have to deal with low-effort bugreps from people looking to show that they are now "expert software engineers" because they can write "go find the bugs in Project X" and an AI will possibly do that.
There are some interesting things being built by AI, but 10x the number of joyful software engineers ain't it.
Not everyone is going to publish their creations, not everyone is on HN, and not everyone on HN is going to shout from the rooftops about how they are in that larger group of newly-enabled people.
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.
But are we actually seeing those results or are people just burning 10x more money?
What I'm not seeing is news promoting how we're suddenly doing 10x better in any aspect of our daily lives, nor 6x or 4x or even 2x better. You'd think if we were suddenly getting 10x more productive, even if it was never published in the news, we'd see something, anything that affected our lives in a positive manner.
We're not.
Or at least, I'm not. And a lot of other people seem to be in the same boat.
I think you're 'putting words in my mouth'. I didn't say that people were becoming 10x better or more productive, just that there are 10x more people being enabled by AI versus programmers who see AI as destroying their jobs.
Doubly so when you consider the history of every previous advance that was supposed to free people from the tyranny of needing a developer to make software, from COBOL forward. Some people indeed made their own applications, while everyone else rejected doing that and preferred to pay someone to do it for them.
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.
> And also that on the other side of that grief is something new and potentially exciting
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.
> 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.
Would you look at a piece of code elegantly using high-level abstractions to model some problem written by a human and go "see, look, code is beautiful"... and then look at the exact same code written by an LLM and say "this is slop"?
> Moreover, it's downright offensive for a purported expert to defer their expertise to a SAAS black box.
With programming, it's just the opposite, though.
Programmers work with abstractions. Understanding the full system down to how atoms work is too much to think of when e.g. writing an API handler. So instead you work with modules and their interfaces.
Pre-LLM, if you need some task done, you wouldn't write it all out yourself if there was some external package which did the work for you.
Of course, when using an LLM, you don't need to surrender your ability to think about the problem and solution.
What is an abstraction? An abstraction is far more than a black box. From a previous comment:
> High-level programming languages use well-defined, limited syntax to encode program behavior in a deterministic way. Compilers translate these directions into programs that should have identical behavior between compiler runs and between different compilers. (If they don't, that's a compiler bug, or else an artifact of UB.) This means you can structurally and provably rely on the high-level programming layer without worrying about what's underneath: the assembly is abstracted away. You also get the massive benefit of being able to tweak and improve the assembly layer independently of the programming language layer, so long as the contract between language and compiler is not violated.
> If AI is a "layer of abstraction," so is working with a bunch of contractors to make your app. The definition is broadened to the point of absurdity. (Solely: "I don't have to think about code anymore," which only touches on the most superficial aspect of an abstraction.)
With a good high-level abstraction, I can mostly think on the level of the abstraction, but I can also dive down through the layers and figure out what's actually going on in each one.
> Would you look at a piece of code elegantly using high-level abstractions to model some problem written by a human and go "see, look, code is beautiful"... and then look at the exact same code written by an LLM and say "this is slop"?
Using is one thing; inventing is another. I have seen no evidence that LLMs have the capacity to discover and leverage new abstractions. To me, that's the most exciting part of programming: finding ways to reify powerful concepts and then turn complex tasks into simple ones, ad infinitum.
The Luddites literally turned to violence when their profession was replaced by better technology. This has been happening for generations to every profession. There is joy in learning new things and developing new skills though.
The Luddites did not resort to violence (unless you count sabotage of machinery as violence), and they were not protesting against better technology. They were protesting against the exploitation of the workers that these machines enabled. Formerly, weaving textiles was a relatively well paid and highly skilled profession, then these machines largely automated the process. This lead to the means of production being owned by the wealthy, employing badly paid and overworked people, to make themselves even richer.
The “violence” was a simple act of sabotage, nobody got hurt except some peace of machinery and the wallets of the rich exploiters.
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.
> Everyone in SWE (who got into it for the love of writing code) is experiencing this.
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.
Yeah the AI companies slurped up all our work and are now selling it back to us in the form of trained LLMs.
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.
When I forecasted open-source is going down because nobody wants to expose himself only to be robbed off credit and authorship, he did not initially understand what the claim was. Two years now.
Indeed. This whole LLM has been a planetary scale theft of labor. They stole our labor and are now leasing it back to us as rent-seeking on the means of production.
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..
“If you want a simple model for predicting the
unemployment rate in the United States over
the next few years, here it is: It will be what
[Alan] Greenspan wants it to be, plus or minus a random
error reflecting the fact that he is not quite God.”
-- Paul Krugman, 1997
Note this does not say "it will decrease if we suddenly invent new productivity technologies".
If you mean publishing open source code, they didn’t steal it. We gave it away. Generosity has always been vulnerable to “exploitation,” but some of apparently understood that risk better than others.
My jumping off point was OSS, but I was actually talking about the entirety of human knowledge being used to make a few people insanely rich while sending everyone else who made the LLM training possible to the bread line.
No, I reject this framing. Copyright violations are orthogonal to labor theft. Labor theft can include copyright violation, but it doesn’t have to. My MIT licensed code may not have been violated (it probably was though; where is the copy of the notice that my license requires?) but my labor was for sure stolen.
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.
The same way that Fitzcarraldo stole the labor of the Amazon Indians when they moved the ship over the mountain in order to profit from the rubber.
Previously when I licensed my code with a liberal license I got something back, credit, contributions, connections, job opportunities, etc. The AI companies stole my labor, then they demanded I pay them, and finally they took my job.
Stealing labor itself isn’t actually a thing. Your wages could be stolen, except - what wages? Maybe the output from your labor could be stolen, but not if nothing material was lost. If you can’t point to a specific intellectual property violation, you have no case.
Did they steal Wikipedia? Wikipedia is still there.
Stealing my labor means that I do all the work, and they take all the profit.
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.
It sounds like you're making a claim to someone else's profits with a very thin rationale for why you had anything to do with it. Basically, claiming something that never was yours and has next to nothing to do with anything you worked on.
The entire concept of copyleft and free software depended on copyright: it isn't "take my code and do what you will", it is "here is a very explicit contract that demands you credit my work and even requires you to contribute back in return for my efforts". And like, maybe a world with no copyright at all is better! But that isn't the world we are in: it is no more legal for you to have an AI read my code and then port it into your codebase (which could involve training as an intermediate step or could merely involve the AI searching for it and loading it into its context) than would be for you to directly read my code and port it into your codebase or even for you to utilize a deterministic tool to read my code and port it into your codebase; you only get to do those things if you retain my license on the result and abide by its terms.
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 depended on copyright as the legal structure to uphold itself, a means to an end, not the end itself. As another comment in this thread nearby says, the point was to support the four fundamental freedoms and ultimately a world without copyright on code at all.
And on the first part about AI reading your code or training from it, it does seem like it's legal now, while you as a human reading and porting is not.
Yeah, I never contributed to FOSS to get credit; I contributed because I wanted something improved which I thought was not very good and thought that sharing the improvements with the world would be good (sharing is caring, after all). From a Free Software perspective, I actually quite like that Generative AI allows people to create the software they want and adapt it to their wants and needs. Free Software is about freedom, after all (free as in free speech). Unfortunate GAI tends to be quite resource intensive, which is bad; but local models are getting better all the time, so hopefully eventually anyone can run a GAI model on their own computer and have it create any and all software they want.
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 freedom to run the program as you wish, for any purpose (freedom 0).
- The freedom to study how the program works, and change it to make it do what you wish (freedom 1). Access to the source code is a precondition for this.
- The freedom to redistribute copies so you can help others (freedom 2).
- The freedom to distribute copies of your modified versions to others (freedom 3). By doing this you can give the whole community a chance to benefit from your changes. Access to the source code is a precondition for this.
Note that "preserve licenses" and "preserve attribution" are very much not in that list. Well, the Four Freedoms are what these AI doohickeys enable by being able to in theory reverse engineer or replicate any code, making all code effectively "public", and to bypass any copyright issues around distribution. The fact that AI blowing away the existence of proprietary code also blows away the licenses of FOSS licensed code is a harmless side effect of getting to the Four Freedoms, which was the Free Software goal all along.
If any of this is surprising, then you didn't understand what you signed up for. The AI monkey's paw has curled and you've got what you asked for, sort of.
And none of those freedoms are guaranteed today for all software but we effectively killed the the mechanism that guaranteed them at least for open source software.
it feels like a transition from artisan level work to factory based industrial manufacturing.
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
And the same thing will happen to devs that happened to factory workers. Jobs will get moved offshore, and salaries will drop, mass layoffs. And SWE's don't have unions like factory workers did.
It's long since past, and it's not like unions would've saved the industry from AI, like Europe has unions too (and their software industry is quite poor compared to the US in general, partially due to higher regulation on companies and workers both).
It doesn't have to be one or the other. Having a machine that does most of the coding for you means you can have more of the collaborative, human input. You can focus on the UX, you can fix all bugs, you can experiment with new ideas faster than ever. There will be zero excuses for poorly done work - think of all the half-assed corporate apps you have had to use during your career. I think this is the start of a great era in software development.
This is the end of software development and the beginning of software management. The problem is you don't need a lot of people to do the later. There are going to be a lot of devs out there vying for the same jobs in the future drastically driving down salaries.
I am inclined to think this way in a commercial setting, but for open source contributors, where their work was done for the love of coding, I can understand feeling like the end has arrived. In my company, I'm building the mechanics of factory-coding and trying to get the team think in higher-level terms: what needs to be built, what does it need to do when it deploys, and how can we build the factory that gets that done repeatedly? To me, building the factory is an interesting problem and agentic coding does not eliminate the higher-level design that absolutely must be done, all of which is, to me, the interesting part of software dev.
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.
I imagine it was what it felt like for miners when they went from working next to each other with pickaxes, to operating massive machines by themselves, although that timeline was much longer and less jarring, taking a generation or two.
I wonder if we will see new platforms emerge around human-only collaboration. There's nothing stopping people from continuing to work on projects in the traditional way. It's just harder on big platforms like GitHub which are designed around public bug reports and pull requests.
I hope so, and not only for code but art and even mindless entertainment. But you really need to draw a very hard line or there will be a line of useful idiots insisting that their "minimal" clanker use is OK and that they should still be accepted in your community. And then the next one will want just a little bit more allowed and before you know it your community is no different from any other on the net.
One of my favorite hobbies used to be to explore GitHub and see what interesting projects were created or updated. I loved to checkout and learn from others.
Everything is weird now.
There is this belief amongst software engineers that humans are always part of the process and that AI prompting is just the new IDE. Yet, every new model release is even more optimized to replace programmers and is pushing us one more step closer to that reality.
Right now (Fall 2026), we focus on empowering agents and LLMs and that's the current challenge facing software. However, once we solve that, the LLM really can kind of take it from there.
I know I'm being dramatic when I say this, but it really is how I'm starting to become jaded from all of this AI tech. It looks more and more like we're becoming a world full of really brilliant humans who are bored because there aren't any intellectual problems left to solve. That is not a utopia. That seems like a dystopian nightmare.
Just like with mathematicians, is it all really just about people losing a random, replaceable job and what they enjoyed about it?
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.
> One wonders whether a generation that demands instant satisfaction of all its needs and instant solution of the world's problems will produce anything of lasting value. Such a generation, even when equipped with the most modern technology, will be essentially primitive — it will stand in awe of nature, and submit to the tutelage of medicine men.
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 point isn't some people not understanding whatever stuff with limited effects, it's no people understanding anything important.
Using an LLM doesn't mean "I must write this code in a way that I cannot understand it". I'd encourage using the LLM to write code in a way that you can understand.
> When corporations own it exclusively, they can do whatever they want with it.
This is an argument in favour of open-weight models.
but so many things we have are black boxes now. do people need to understand how cars work to drive them properly?
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).
The difference is that people who make the cars and the mechanics who repair them still understand how that black box works.
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?
probably not, but my main gripe would be that there would be no way to punish anyone if something went wrong with the car, not (necessarily) that human judgement or expertise is still needed as part of the car design process.
So yes, maybe we will all just exist to get sued at the end of the day.
> What happens when you don't understand them anymore?
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).
I used to love making the occasional event poster for friends and signs for local businesses as an aside to my day-to-day work. Early in my career I was in advertising and design, and still love my art.
But in the last year you just don't get these jobs anymore, due to AI. It's everywhere and I totally understand the feeling uncreative people get when they are finally given a tool that lets them create something.
What makes me depressed however, is seeing that the vast majority of people are perfectly fine with AI slop posters etc... Highlighting to me that they likely didn't actually value or appreciate the good design to begin with. A lot of people don't care about the human element to anything.
> A lot of people don't care about the human element to anything.
I think this is a crucial point. The AI era seems to bring out two distinct movements in people: Those that always saw the human as a necessary bottle-neck, and those that actively sought and enjoyed interacting with another human brain.
> Those that always saw the human as a necessary bottle-neck, and those that actively sought and enjoyed interacting with another human brain.
Yep. The problem I feel is that humanity and society evolved via human interaction over the span of tens of thousands of years. And then in the space of a decade some of the people who benefited from that development feel so comfortable in their lives that they seek to literally end that method of progression for our species.
Yeah the most depressing part of the AIpocalypse is the realization that most people simply don't have any taste whatsoever. At least that's the only way I can explain why these horrid slop posters are thriving instead of essentially killing whatever business is putting them on their windows. It's not even just the missing human element (but it is that as well) but the level of quality that the average person seems to be OK with is just so low. Lower than what pretty much everyone should be able to make by hand.
I never understood this. It's like home cooking, I don't stop just because a professional chef can make it better. Seems like many OSS contributors wanted to have their cake and eat it too, they're not really contributing for the love of the game but for external factors like credit or admiration or clout.
Would you keep cooking if a Star Trek style replicator existed that could more or less cook anything in a minute? What if all of your friends and family were satisfied with the replicator food and weren’t willing to wait for you to cook anymore?
It's like ordering food from outside versus cooking food yourself. Yes, you could order it and it might be appreciated but cooking something for others will, if they're good people, have them definitely appreciate your effort. And there's a difference between cooking for sustenance daily, in which case I too would use the replicator just to save myself the effort, and cooking for fun and special occasions.
This is analogous to coding as a profession versus for fun. For work I'd use all the tools at my disposal to get the job done because that's what matters in a corporate environment. And for home, well, I'll code some things and vibe code others, depending on how I feel about a certain project. I think OSS maintainers lean too hard in setting themselves and others too high of expectations such that they burn out like this, many such cases, when in reality they shouldn't take it so seriously. Work on your code when you want to and leave it alone when you don't, the license literally says "as-is," so follow that philosophy.
I am always doing everything alone, so I am liking the LLMs a lot. My own code is very good, but nobody but myself cared about that. Even the one job that I got which allowed me to gather the capital needed to move into trading feels like complete luck. This is a new golden age for entrepreneurs and apocalypse for wage workers. Nobody whose status depends on his own productivity would complain about it going up sharply.
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.
I think you’re heavily miscalibrated. Anyone can use LLMs, the floor has risen drastically. What supposed productivity you achieve, every one else can because there’s 0 skill to using LLMs, this is probably a much darker future ahead for you than the rest, you’ll have to keep constantly agent maxing to try and outcompete millions of other humans doing the same.
I'm not sure this is credible. Using AI very much seems to be a skill with its own craft and best-practices. I could imagine a similar argument was made by painters about photographers. Yet, we believe good photography takes skill.
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.
People are afraid to do work on cars (and mains wiring, and plumbing) because even though the work is relatively simple the consequences for fucking up are so severe (severe damage, injury, death, etc). That's most of the reason they pay someone else to do it.
For the vast majority of software that isn't the case at all.
Most people don't understand computers and do not even know what is dangerous and what is not. They won't suddenly become vibe-programmers lest "AI deletes all of the personal photos and will send them somewhere".
No need, I still have vibe coded stuff from the times before monthly plans for Claude and Codex over a year ago. I can jump into most of them and just start adding stuff without any major issues - they have better test coverage and CI than many projects at my day job :)
It's surprisingly easy to do that shit properly when budgets and time aren't an issue.
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).
LLMs are constantly driving off the cliff. Constantly. They require some actual agency steering the ship.
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.
lol wat? there is a very clear gap in output between what a good swe can produce with ai versus a bad one, let alone a random person who doesn't even know what javascript is.
One thing that strikes me about what is going on: we have been promised this renaissance in software, yet, it been 2-3y of accelerated agentic/LLM development and I haven't see anything great coming out of the LLM developed software world that is worth looking at. Let's take open source as an example: No new 3d/2d CAD packages that are more capable than Blender/FreeCad/KiCad, nor Blender/FreeCad all of the sudden became much more capable than they were just few years ago. I also want to point out,that I haven't seen any new LLMed software in Electronic/ElectroDynamic simulation, something I would LLM to excel at, beyond what already existed before.
No new software that is equal to Adobe Creative Suite (that Vibed coded replacement is trash IMO), No new Operating Systems that are taking over. No new utilities that are super useful in day work. We are using the same software, we've used before LLM arrived.
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.
One of the things I've started noticing at work is that for the last few months I was banging away at having multiple streams at a time. But over time, I've reached the point where there is some human blocker on the stream, be it QA to test it, other humans to have attention to pick it up for the next step, deployment of some sort, or in some happy cases, it made it through to deployment and the problem is now solved. In the past couple of weeks my multistreaming is toning down and more often I'm back to one stream of work. Often a much higher level stream than I would have been working at a couple of years ago, and I'm moving faster in a number of ways. Some of that "faster" has been burned up in things like changing environments that I previously just wouldn't have done, though.
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.
I think 2x or 3x is maybe too generous. Obviously pure code output is extremely high but I would think for something to truly be a 2x a company would either have to be able to effectively get rid of half its SWE workforce or produce enough to make some notable increase in revenue.
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.
We might be producing shovelfuls of code at 10-20X the previous rate, but we're not producing value proportionally, maybe we are even inversely proportionally.
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.
People say this but I mean, don't you still want that 5 to 15 percent of time back? I certainly appreciate my time back with my LLM work, but then again I work remotely so there are lots of benefits to extra time around the house while the LLM churns away.
It depends on the role too. I had jobs where I was in meetings all day and others where it was all asynchronous and had essentially no meetings, and so in those latter cases I could get a lot of time back.
> Let's take open source as an example: No new 3d/2d CAD packages that are more capable than Blender/FreeCad/KiCad, nor Blender/FreeCad all of the sudden became much more capable than they were just few years ago.
I know nothing of Blender but in the FreeCAD community there was a little disagreement about use of AI and it has taken some time to find a policy most people don't mind (one of the original authors stepped back from maintaining because he disagrees with this policy). Plus they have also traditionally had quite a long release cycle.
Nevertheless, FreeCAD is much more capable than it was just few years ago (largely through slogged human effort in the main) and the velocity of change is very evidently increasing with AI.
However, CAD applications are quite a difficult example to pick, because bRep kernels are hard, FreeCAD is not in control of theirs (it's OpenCascade, also used by PrusaSlicer/Orca Slicer for STEP conversion), etc.
AFAIK OpenCascade's development has picked up recently; whether they are using AI I don't know, but I expect they are not accepting AI commits from outside sources.
I have. I use a MacBook and now lots of games that didn't have native ports now do, via vibe coding and AI reverse engineering. This would have taken years before and now anyone can ask their LLM to do the same for their favorite game or app to run on whatever device they want. You say the vibe coded replacement is trash but in some time it'll because better simply because users will iterate on it adding features that they want, not what a corporation thinks they should have.
OP of the top reply. As I have mentioned, RE process is pretty much automated thanks to AI. But once again, RE existing software/games and not the same as developing a new software/games. AI being great at RE is an inherent property of AI itself.
I think this comment is premature. It hasn't really been 2-3 years accelerated LLM development. Even 6 months ago, LLM development was much less refined than it is now. There's been a few years of figuring out how to use these tools (which are constantly changing) and maybe a year of using them effectively.
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?
>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.
My projects are so niche that almost nobody would ever care... but think that might be the point. Maybe there aren't any decent MS Office or Photoshop clones but there's also a bunch of really interesting niche software being created nobody else would ever care about.
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.
I worked for mid size Insure tech company that has good footprint in the industry and two years ago, LLM were good enough for us to refactor & rewrite major tech debt. Even two years ago, models were already impressive & capable. For me, the timeline been two years and counting.
A year ago I was also doing major refactoring and it saved several months of time (probably wouldn't have even attempted it otherwise). But LLMs still weren't as capable as they are now for more interesting tasks than that. I struggled with an LLM assisted conversion project, eventually just shelved it, and started it again recently and the difference is quite noticeable.
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.
I thinks what's more happening is bespoke software made by people for themselves mostly, often with a user count of 1. That's quite a huge change, for users of that software. Of course such things are not replacing existing apps.
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.
The quality of everything has gone downhill -- software is just part of that trend. Everything is about extracting as much value as possible for as little cost as possible. We've optimized everything, including software development, to that. Quality requires a level of care that society is no longer able to achieve.
I recently re-read Venus, Inc. and the number of times that the satire didn't hit because it read too close to our current reality was quite depressing.
People say that, but it doesn't hold up. We spend more money than ever before on software development. Just not on quality. There is something else going on here.
The last time I encountered a bug was in some open source software. I asked AI for help. The AI read the 350,000 line project on GitHub and said "yes, that is a bug, but nobody has reported it", and then showed me the code related to the bug.
I cloned it and fixed it myself.
Similarly, I've asked questions about open source projects with poor documentation. The AI just reads the entire source code and finds the answer.
It will be interesting to see if having access to the source code becomes more valuable because of this.
Yes, I think having access to the source code is increasing going to be a requirement for enterprise software... and I think that's going to come from security rather than bugs.
The problem with self bug-fix is that it leaves you on a unique fork. Repeat that 100x for 100 different users, and now how does upstream add a new feature reliably? Everyone's going to merge into their own fork?
My guess is we'll still be using upstream-authoritative, but at some point user-initiated patches will flood over the dam and become the new normal even in formerly closed-source enterprise.
Because source is going to be immaterial (decomp will be good enough) and the security benefits of customers being able to self-scan are pretty high, especially in an age where time-to-exploit approaches zero.
Yes! It's just repeatedly asserted that vibecoding is taking over, going to replace us all. I have yet to see evidence of it (no, marketing stunts aren't evidence).
We have seen a huge proliferation of vibed audio plugins. Most of them are very bad or replicate things DAWs can already do. And vibed video games. If anything it has become harder for many audiences to find good software because algorithmic curation is driven by gameable engagement metrics not quality.
Hmm, that may well be true but I don't model with any of those packages. I model with `build123d` or `openscad` using Claude Code to do the heavy lifting. I describe what I want, measure and make up ideas and learn a little terminology (fillet, chamfer) and then the machine builds me the concept that I then print and deploy into action[0].
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".
I agree with this to a large extent. This new technology is lowering the barrier to meaningfully work with existing tools. It's increasing the magnitude of jostling that may very well bump us out of a local maximum sooner than before.
I've recently been getting into musics an synthesizers and VSTs.
If you can write a program that produces a convincing piano sound people will pay you $100 per license, or more. Same thing for a violin. All you have to do is make it conform to the VST API specification and provide a bare bones UI.
But, where are they? Where are the new VSTs? I haven't seen any. Not at an accelerated rate at least.
If you can write a new and better geometry kernel for CAD software people would be willing to pay thousands of dollars. You're not building the next Google with massive distributed systems here. It's just math. To build a geometry kernal you just have to do the right calculations on a local CPU, easy. But where are the new geometry kernels? Who is vibe coding a geometry kernel?
Truck is getting a lot more work done because of AI, I believe.
But geometry kernels are severely difficult work, the maths is hard, the calculations are hard, efficiency is hard, and also policy is hard; decisions around topological naming, around the correctness of fillets (which major CAD packages do not agree on), etc.
They are among the most difficult things people have ever done with computers, which is why there are only about four of them, and they do actually put assembling large distributed platforms in the shade.
Now, will LLMs be able to help with that maths? I hope so.
Re: VSTs, I don't know much about them. The pace of development of AUv3 plugins seems to have increased, though.
I personally think the agents and the harnesses themselves are pretty cool software. Claude code, openclaw, hermes, and now the consumer facing ones. I have been doing all kinds of cool things with them. If you gave somebody the claude desktop app with blender mcp 10 years ago their minds would be absolutely blown.
A lot of these harnesses and mcps themselves are built with llms.
No it’s not. There’s nothing technically impressive about this.
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.
I use LLMs but this is probably the worst example to bring up, because there is nothing special about this as another Linux distro versus actually interesting projects like decomps and recomps.
We're pretty much still building the tools to build the software renaissance.
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".
I thought about doing that with Todoist.. and then just decided I'd rather pay the $4-5/mo or whatever to not have to use unpolished apps and sync server
I keep paying for Obsidian for the same reason. It's central to the way I work and has an ecosystem of stuff-that-understands-Obsidian making it useful.
But, for example, I did replace IRCCloud with my own custom setup. Been daily driving it for ... three months now? Took me two to three evenings to get an MVP and 3 months of iterating while using it to get it Just The Way I Like It.
I was also a paying customer of Hazel[0] for two major versions, I didn't use the majority of features so I made my own that does the things from Hazel I needed plus a few extra things that nobody else is going to care about but make my life easier.
All of these are just for me, they are open source with a permissive license, but I have zero intention of making a "community" for any of them or trying to make them into commercial offerings.
They do what I want and if someone likes them as-is, they're free to use them. If they need something else - that's why we can fork repositories =)
That's funny, I felt the exact opposite and found it fun to research how to make things sync properly and architect the solution for my LLM to then build. Currently I'm using Loro for the syncing.
Maybe it's not better. Maybe your vibe coded solution is much better for your specific use case. For example I recently wanted something that could track my watch history for shows I have locally and sync them to all devices, sort of like Jellyfin, and I couldn't find something that did what I needed so I had to make my own.
Depends on the price. There are many applications that people and small/medium companies pay hundreds of euros a month for and only use a few features.
At that scale there are cases where it's worth it to create an AI assisted bespoke version. I've done two of these at $dayjob and a bunch for my own purposes. Either replacing subscription software or highly specialized versions of pay-once software.
Perhaps BigCorp is only just starting to use agentic dev properly. Plus they’ll have loads of bureaucracy still to get code out of the door. Might take another year or two.
Very simple - you don't see it because it does not get released. What you see get released, though, most people would not register so fast on their radars. And it is very bold to say "no, I don't see how antirez is puttin LLMs to work" or something along the line. Well, very often you don't see most of such people's work until it's literally everywhere. Llama.cpp a good example of such software, and yes, it accepts LLM-assisted code.
> Reverse Engineering became much easier, it is plus. Finding vulnerabilities also became much easier, another plus. Models are getting better.
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.
I don't get it. Like obviously any in development software will have bugs and missing features, and when it's replacing something as complex as Photoshop you can expect more issues compared to a simple text editor.
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.
> No new utilities that are super useful in day work.
thats the renaissance for you - you no longer need the utilities, just ask the agent to do that, whatever it might be.
building great software has not been constrained on throughput nor technical for a while. it's limited by the human aspect, and there is no reason imo to believe by default that ai would change this.
I don't want to toot my own horn but I developed a mix of Photoshop / Pixelmator for AI assisted Photo editing and I use it every day instead of Pixelmator Pro [ 1 ]. I think it is in many respects the better app because I added many features I always wanted to have (replicate layers, any content as masks, crazy effects, etc). Granted it is buggy and still needs polish, but most software has bugs. So I've completely switched over to a complex AI generated 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.
This is not representative of what's actually going on, though. It's much more like an artisan carpenter who wants to build houses using bespoke hand-tools and precise joinery. That's great, and there are beautiful works of art created by those skills. But in the last year robots were developed that build homes exactly to spec. The robots work 24/7 and finish entire houses in a few days. That same skilled carpenter is being hired to monitor the robots and check over their work. There is, economically speaking, no place in the process for him to apply his craft. In the time he rips a single board, the robots build the wall. In the time he drives a handful of nails, the drywall is up and the paint is drying. He costs $200/hr, the robots cost $1/hr.
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.
This analogy hinges on the economic argument (that a robot costs 200x less than a human), but far from being an established fact, that notion is more or less entirely based on vague, misleading and exaggerated claims from the token peddlers.
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.
> Its extremely naive, though, to simply buy into the PR from the peddlers
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.
I think the OP was talking about robotics specifically and not tasks like the one you described. Nobody can argue that for tasks like your binder replacement cost savings are enormous.
No, they were carrying on the analogy about artisanal home builders being replaced by mythical robots that do it perfectly.
In this analogy, we human software devs are the artisanal homebuilders and the fancy LLMs are the robots.
And yes, there's quite a cost to build all the datacenters for all the AI that Big Tech believes that people will want. But as you mention, the cost savings are enormous for even despite that, and will make a much larger set of problems accessible to real software solutions, that previously either stayed on paper, in peoples' heads, or maybe an Excel sheet somewhere.
Almost anyone will accept a 50% quality decrease for a 99% cost decrease. The only reasons I'm not living in a stack of cardboard boxes for $1 rent are that it's illegal, and the police won't protect it from break-ins.
With the money you wouldn't be paying in rent, you could build a heck of a lot of stuff. Cardboard box is a metaphorical starting point. You could build some kind of little wooden box vehicle and travel the country. If you stayed in town you could have takeout for every meal too and still have money left over. The point was not paying rent to the corrupt system, the point wasn't the exact living specifications.
You can go and live in cardboard boxes right now for free. Why aren't you? Being technically legal or not is besides the point - there are places where it is tolerated.
>You can go and live in cardboard boxes right now for free. Why aren't you?
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.
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.
Not that tired old parable again. It doesn't hold.
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.
> But in the last year robots were developed that build homes exactly to spec.
That is not what's happened at all. That's the entire point of the discussion. The robots aren't as good as humans, they can't follow specs or instuctions as well, and their output is markedly worse than human output in a lot of ways.
The only difference is that the robots are faster and don't need rest. And we're discovering now that that is not good enough.
Unless they manage regulatory capture and with this kind of money at stake I would not bet against that. It's going to get very ugly. Read up on transhumanism to see what the stakes are, they're playing for keeps here.
Regulatory capture - I tried to bring this up in another comment of mine, but not much resonance.
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?
> How would regulatory capture of LLM realistically look like? Do you have any ideas here?
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.
> Read up on transhumanism to see what the stakes are...
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.
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.
Work, sickness, and death are -for the most part- all things that are awful. As it turns out, there are many more than three awful aspects to being a present-day human. For brevity's sake, I picked three that seemed to me to be most likely to be widely understood and widely considered to be awful.
> 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_.".
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.
Except the robots actually cost far more than $1/hour, it’s just that the owners have enough capital to operate at a massive loss until the carpenters have largely been replaced. Only then will we learn what the robots actually cost.
Plus robots (software and hardware) will increasingly be used to invade even the most basic privacy, to control people etc. They are not just productivity tools anymore, they are also mass surveillance tools and every bad thing that comes with it.
Self-hosted and/or open models are fast approaching (or have reached) the quality bar necessary to handle a great deal of software development work. The models needed to cause the OP's problem are becoming commoditized.
Forget robots, what you're describing already happened in the trades. Modern construction techniques are all about lowering the skill required to build a house and standardizing the process so it can be done with minimal training. E.g. drywall and caulk replacing plaster and precise joinery.
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.
I have been told it is the future of sleeping, and if I don‘t sleep with a plastic cover directing the pooping agent I will fall behind and eventually become obsolete. And that sleeping like this will eventually cure cancer and build swarms of mining rigs in space.
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.
100%. Although large handwritten software code bases are often indecipherable to the average human being (actually, almost always indecipherable to the average human being, but often indecipherable even to skilled software engineers too); sometimes due to inherent complexities, sometimes due to Enterprise-isms.
Those slippery slope arguments people make assume that once a technology introduces a specific risk, society lacks the agency to create counter-measures, new norms, or alternative uses for that technology.
History shows that when a slope appears, regulation steps in, technology evolves to solve the problem, or the culture shifts to reinterpret the tool.
In almost every case, the feared bottom of the slope was never reached because humans constantly built ramps or bridges along the way.
sounds like bullshit. I'm actually glad that the people that hold this view are getting out of the way, they are the people that gatekept open source and made it annoying to work with. Do I miss having to wait for months to fix a trivial bug because I first need to convince 15 people that the fix is even necessary?
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.
I have a strange feeling that we'll see an increasing number of these to the point where we'll start to see legitimate issues in ecosystems as real talent exits (due to projects no longer evolving, being deprecated, etc as a response to AI).
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.
Well, people aren’t coding to keep their skills and knowledge fresh.
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?
I tried to write a little project the way an AI would, going gung-ho and not stopping to ask if this is a good design. It actually went okay and now I have a silly thing that wouldn't exist otherwise. It was still slower than an AI, but only took a few days and now I know how the thing works inside, which I wouldn't if I used AI. I don't think programmers can't match AI, I think we don't want the associated responsibility of knowing we might be building a pile of shit. I think we want to avoid writing a pile of shit so we slow way the heck down.
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.
I think some people have a tendency to forget that you can half-ass things because some things only actually need half of an ass (actually a surprising number of things, though it can be difficult to predict that completely accurately!).
I've yet to see a well-loved widely-used vibe-coded (rather than just AI-assisted) app that has gone live, and then gone through multiple feature additions and remained secure and used.
This is exactly how people were told to build MVPs for their startups, I'd have expected HN, being on YC, of all places to understand this, so it's good that you're keeping the culture alive.
> 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?
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].
It's already started imo. Spotify's desktop app on MacOS is becoming unusable (up to ~30s delays after clicking a button before the action runs—used to be snappy). I just saw the Epic Games launcher thing. Anecdotally, countless websites just break at random (more so than even when it was 100% human).
Spotify doesn't care about how fast their app is because for a lot of their users there is no alternative. Everything else is paid only (Apple), has a bad quality library with duplicates and trash metadata (YT) or is a lot less convenient (physical media and most pirate options).
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.
Presumably, private companies and individuals with the money to pay for tokens continue forward rolling their own in-house solutions and slowly abandon OSS as it rots. What happens when each organization has its own entire stack though? Sure, FAANGs already were like that but most smaller companies were sort of "speaking the same language" and reusing software like apache, mongodb, nginx etc.
> What happens when each organization has its own entire stack though?
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 have a different sentiment than most of what is shared here. I do enjoy coding. I do enjoy solving problems, but I feel like most of the work in my corporate/startup job life always have quite a significant amount repetitive and boring work (e.g. updating library for security reason, adding new CRUD endpoints, updating the app as per latest user study, etc). And so now I feel liberated because I can focus on problem solving instead of those tedious work. Given a problem, I can focus on thinking how to solve the problem instead of spending time coding up the solution. I don't blindly tell AI the problem and have it come up with a solution. Instead, I am particular in how I want problem to be solved, and then I will tell AI to implement it. And so in a way, now I have a 24/7 junior engineers that would get things done very quickly according to how I want the work to be done.
It's true that my ability to code would have atrophied, but I think it would happen anyway when one gets promoted and have to spend less and less time coding and more architecture or reviewing their juniors' work.
I still review the code created by AI very closely, so I don't feel like my ability to code degrades as much. Yeah, and when I feel like I have time, I would like to work on the code myself. Plus, if I want to, I can always code things by hand in my own personal project, for example.
One of the most underrated abilities in this industry is knowing when to walk away from something that no longer bring you joy. It's sad that it has come to this, but I'm sure the OP will find something more fun to do with their time.
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 have wondered where we are with stuff like this.
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.
Eventually maintainers will just throw up their hands and leave the repo under the care of a LLM. "Merge PRs that align with the product's vision/design/intention. Meticulously examine for malicious contributions. Email me if there are doubts."
The top tier LLMs would be as good at this as a lot of maintainers, right?
> 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.
yeah maybe. i think it's become a lot more feasible to fork and modify vs wait for new features or fix bugs. the future might be that the main repo for a project forms the "core", and consumers use their own forks which are automatically kept up to date. the core doesn't need to be as configurable or flexible as the forks can add functionality directly. in theory it could be easier for the maintainers.
I feel sad about this passing of the age of programmer because programming was a good way to make it big when you have nothing. You start your life in a poor family, you fall in love with technology, study hard and make a good life for yourself and your family by becoming a programmer. Now this career avenue (like many others) is closing because of AI. Everyone becomes less valuable and less happy.
The source is open, but the ownership of the "text" and 130 other gems still resides with the blog author... maybe he could add the projects to https://www.codeshelter.co/ or something.
It's definitely a great language to have fun with. I enjoyed writing the book. Moreover, I wrote it for my son who was 5 years old back then, in hopes it will be useful over time, pretty much like books about bash/sh were useful for decades.
I still remember those days I was working on my book to and from work. Those memories and dreams are priceless.
I haven't updated the book intentionally to keep the "Last updated on 2022-05-26". Was it the last Ruby book written without AI?
I can't speak to Ruby or other programming languages, but as a creator and occasional hacker with a media background, I have found AI to be incredibly liberating.
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.
> 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.
That factory produced chair is way more likely to end up as landfill than a hand-made one. The AI-generated slopware generates the landfill (ie pollution and the jobkilling) before it's delivered to you.
You can completely clone board games and steal from their original makers OR you can make something new that didn’t exist before because there’s tools to facilitate your work that you didn’t have until now
Transcription is one legitimate use for LLMs that archivists are excited about. It's a case of a tedious task that frees them up to do the more interesting parts of the jobs, and the product is something that can be easily compared to the original and verified. This is very different from something like writing prose or code.
It definitely varies person-by-person. People who enjoyed l33tcoding, the act of programming, and the building process feel one way vs. the people who are in it for the end product, or enjoy the design process feel another. In a way, it's almost like those who enjoy the journey vs. the result (although the design process is also a journey).
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.
One thing I always like about coding is that when something is annoying or hard to do, you can improve things so that it becomes easier and more straightforward, at least in principle. For example if a library is badly designed, buggy, or just doesn’t exist, you can write your own. If tooling is bad, you can improve it or create your own tooling. At the extreme, you can create your own programming language, database or operating system. In the small, you can write your own macros and wrapper functions.
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.
Actually nobody is preventing you from building library with LLM.
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.
I think you missed my point. Yes I can work on a library with an LLM, but I can’t work on the LLM itself, which is the interface I’m corresponding with now, where previously it was the library. Previously, the thing I was interacting with was malleable in a highly well-defined way. Now it isn’t.
Yeah: in a world where I am coding, I can put in the effort to make the software work the way I think... to use an LLM, I largely have to think the way it wants to work / converse.
I get it but not every editor/IDE was vim/emacs. Some tools were more or less rigid and today you still can have malleable workflows through custom workflows and harnesses.
It's not like you need to limit yourself to Anthropic and Claude Code. I myself developed "rik" a harness that can modify only a single file and hook up blazingly fast models for it to be like wide-context autocompletion.
1) I really wasn't talking about an editor at all: I was talking about the libraries and frameworks you use to program, so that the way you interact with the problem space is the way you think. I mean, even as someone who has almost exclusively used vim for the past 25 years of coding, if I went back to the prior 10 years before THAT of using Visual Studio, I'd have described this no differently.
2) In fact I have never used off-the-shelf coding assistants (Claude Code, etc.) and have only used my own harness; yet to me the issue is that the model itself often likes to think about problems in ways I find frustrating and unhelpful. The issue isn't at the level of how diffs are applied (that's commodity) or how context is managed (I can fix that) but in how problems are described, how abstractions are managed, and the language used to discuss them.
So many people here are talking about not coding anymore. At the same time, when a project is coded by AI, it's mocked here as being vibe-coded, and many people say they won't use it.
There seems to be an inconsistency between those facts that would be worth discussing. I'd be interested to hear anyone's thoughts.
How do you resolve it? Do you think people are creating software they wouldn't use, perhaps because they are forced to create what they would judge as crappy AI software on their jobs?
Ah... maybe I misstated my question. There are people who are stopping working on FOSS projects because that's no longer fun. But what I meant to focus on is people who say they are still creating software, using AI, and therefore miss coding. And the fact that there are people who say they wouldn't use a vibe-coded project.
If those are overlapping groups, there would seem to be an inconsistency, but maybe it's resolved as I mentioned in the first paragraph of this comment.
Some people are producing vibe coded garbage with LLMs and some people are protesting and getting out of the business. It doesn't have to be the same people.
But yeah, people who are doing both could be forced to use LLMs at work and unable to leave. And when those LLMs inevitably fail because the codebase is too complex they have to refactor with LLMs because what they originally crapped out is incomprehensible and no one bothered to read and understand it while it was being developed. I expect a lot of people don't like that we are becoming dependent on funnelling money to a third party to maintain the crappy software they enable to begin with.
I think the entire industry is about to have another recognition about the importance of quality in software development. Just like they did with lines of code as a metric, and outsourcing with minimal communication.
I'm so conflicted about this. I do love being able to create and bring ideas to life at the light of speed. But I also see the unique value in the experience of coding, deeply caring about the craft. Can we have both?
> I'm so conflicted about this. I do love being able to create and bring ideas to life at the light of speed. But I also see the unique value in the experience of coding, deeply caring about the craft. Can we have both?
I'm working two entry level jobs, and I have spent maybe upwards of 100B tokens in the last year? on a project which isn't more than 250,000 lines of code. I have done countless refactors and care a lot about the craft.
Performance is a consistent issue. Ive had to build a lot of optimization and benchmarking tools in order to keep things functional. Every couple of weeks I have to refactor everything to keep load times good. Both on the database side and on the lisp side. The models, by default, will happily build functions with 26 database calls or millions of recursive steps.
More frustrating, though, was probably the 3-4 weeks when i migrated the storage concept to "one big tree". This was an aesthetic thing that I later found out some of the old lisp machines also did. But - I didn't think about it initially when I built it. And switching things at that low level was difficult. There was a lot of legacy data stores along the way and programs were using the old store constantly. Very irritating. I still wonder if it would have been faster to just completely rebuild at that point.
An ambitious user could probably run it themselves and have it function, or I think I could run it as a managed service for somebody and have it be good, but the UX is still not really ready for primetime
I think it's like any other craft. Buildings and most furniture are built using mass produced lumber and machinery, but many people still enjoy doing woodworking by hand.
Yes, but it may be hard to justify the "experience of coding" from a commercial standpoint.
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.
"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."
Well, it might be a rational view for the author, but for me that's like saying "I no longer like cooking, because there are companies providing the recipes I like as ready-made meals", or "I no longer play my instruments and make music, because there are machines which can create the music I like"; both topic apply to me personally, and I wouldn't agree to the statements. I still like to design and implement systems, make clever architectures and solve tricky problems. Why should I care if other people use LLMs to do so?
For most of us, as much as we enjoy making clever things, without the context of a job there aren't that many fun problems we have time to engage with.
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.
I'm still in business myself and I see the problems. But the article was not about what you suggest. It was about contributing open-source to a community and whether it is still worthwhile and fun, even if some community exponents started to use LLMs.
If AI can do what we deem creative work, say implementing a chunk of a Ruby gem, should we still consider such work creative in the first place? I was wondering if we should face an uncomfortable realization: much of what we call being creative was really the cost of overcoming cognitive friction due to the limit of our brains.
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.
> If AI can do what we deem creative work, say implementing a chunk of a Ruby gem, should we still consider such work creative in the first place?
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.
I think this is really an excellent comment because it challenges us to reframe what we take for granted as creativity, enforces us to reconsider what creativity could be in the face of really powerful automation.
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.
This is going to be happening all over the place. I've released some code in the past but I'm definitely not going to do so again without ironclad agreements in place about AI ingestion. I'm not going to help the likes of Amodei and Altman to a further increase in their power, my goal was to help normal people not weird cult-like corporations and I suspect that a lot of people that have released code over the years have the same attitude (though obviously not all of them).
Yes when it becomes clear that half the participants aren't playing by the same rules you were its only fair to flip the table and walk away.
I'm no longer motivated to share when that only ends up feeding the machine intent on replacing me and I'm no longer interested in looking at other's contributions when it likely contains vomit from that very same machine, whether disclosed or not. This doesn't bode well for the future of open source or free culture in general and I do hope that well get the societal and legislative changes needed to make sharing a remotely fair proposition again but, considering that the entire economy has been essentially a big casino for over a decade now and all the winners are not held accountable for the issues they cause because doing so would bring the entire system crashing down, I'm not optimistic about that happening within my lifetime.
> You want changes? Go ask your agent or something. I’m told that’s what you’re supposed to do now.
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?
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.
I am pretty sure that the whole open source ideology was developed and disseminated exactly by a cartel of those massive corporations with the sole intention to amass as much human-written code as possible to train AI.
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 really would like to believe that GPL is the answer here but I know it won’t change anything for them. If they can steal and destroy books, nothing will ever stop them.
We need to go beyond GPL and have the government invest in public software programs.
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.
Paying for something that has been scraped, trained on hundreds of billions of lines of code (mixed up GPL and other licenses) and being resold back to you is trickle down?
Paying 20$ per month to double my productivity, answer many of my health questions with good accuracy, help me with DIY projects around the house and do almost as good of a job as a psychotherapist is trickle down for me.
And you don't even have to pay - there are perfectly fine free tiers and open weight models.
It would not be absurd: Ruby (not Rails, Ruby, I insist) puts very high the "joy of coding", its syntax pleasant to write and read for humans, and its welcoming community. It never was the most performant, it's been a long time since it last was the cool kid on the block, but all those other aspects make it very attractive for very different reasons. That's a big part of why I started writing Ruby a long time ago vs. other languages, and why I enjoyed it so much, and I don't think I'm the only one.
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.
Ruby is definitely not bringing me joy since I have to see needlessly depressing articles like this. Rails did not bring me joy 12 years ago when I needed to update dependencies for a legacy cloud service we relied on. Ruby's middling syntax didn't bring me joy either, despite the hype of the time. My child named our cat Ruby, and I had to explain to fellow nerds that it wasn't inspired by the language.
I find it amusing to hear these anti-dhh rants. These old schoolers should remember the old “Rails is a Ghetto” rant. It’s got much of the same technical criticism in there. DHH was never a star, but he’s good at marketing.
I really enjoy Ruby, rails, and agent-driven development. So I’m probably not the target audience. To each their own.
The robots aren't as good as you think. Its just that we're sceamed at all fay by hoardes of people saying they are.
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.
But the robots are mostly "fine", and "fine" is good enough for most things we get paid for: enterprise SaaS products, mobile apps, etc.
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.
This. Enterprise software is going to get a lot worse.
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.
That's a fear I have, yes. Not much we can do about it, though. We've collectively been pretty lucky so far, in terms of compensation. Can't last forever, one way or another.
I know of at least 3 significant rails projects/donors that are actively discussing AI-lead refactors since DHH's talk declaring rails dead. Even if you like the language it's going to be a total mess in a year or two.
I'd rather tell LLM to rewrite project in Python. There is no good reason to use Ruby/Rails now, the community and job market is shrinking and the language is nothing special.
"But even beyond Rails, the Ruby language itself is now substantially developed using LLM agents. Matz is giving talks repositioning Ruby as a “Programming Language for AI age”, to be emitted by LLM agents, and stating that to be able to do that LLMs need to consume more good human-written Ruby code. You might already have seen the problem here."
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.
Why would it be wrong to call Linus a sellout? The guy started with a hobby project that eventually became big, was mostly taken over by corporations, and he is now doing mostly corporate friendly things.
Seems unnecessarily harsh. LLMs are a power tool. I'm not on-board with how it's being used in a lot of places, once things shake out I believe it's going to be net positive.
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.
a LOT of these arguments against New Thing seem to always fall into the category of "I no longer receive social credit for being a productive member of a niche in-group, so I quit.", masked as "They peed in the pool!".
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.
The social aspect of a craft is more than just "clout" or "smart creds" from an in-group, it is also collaboration, commentary, discussion, and the joy of helping others find their feet, for example. LLM-centric programming takes away not just the "clout" part but also all the other less egocentric social aspects of programming as a hobby. If the hobby's ecosystem gets whittled down to posting a finished demo and saying "look at what Claude did," well, what is there to talk about, really? The whole "process" aspect of the hobby gets obliterated, or at least effectively ceases to exist as a social phenomenon.
This is an overreaction. Yes, with AI we can all write our own features and bug fixes for open source, but there’s still community benefit in having a canonical version somewhere, rather than everyone writing bespoke solutions for themselves.
> there’s still community benefit in having a canonical version somewhere, rather than everyone writing bespoke solutions for themselves.
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
I'm not really advocating for any position here - just pointing out that open-source maintainer was a thankless role before LLMs, and it's only become worse since.
I was chatting with a friend who commits on a very well known open source project (~80K stars). He is disillusioned not because the robots aren't good at what they do -- they totally are. The sad thing is that he enjoyed writing code and working with the community on improvements because it was a very human, collaborative thing to do. So, while the robots are great at pushing the project forward even faster and doing fantastic things for security and performance, the joy is gone.
I can totally identify with that feeling. It must be how professional mathematicians are feeling as well.
Everyone in SWE (who got into it for the love of writing code) is experiencing this. 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.
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.
Like many, I became a SWE because I loved writing code. Ruby, in particular (whatever one thinks of its other strengths and weaknesses) can be very expressive and beautiful.
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.
This is super great. God, I wish all leaders had half this much empathy! Good on you.I am one of those weirdos that found joy in business software. The clear specifications, reasonable timelines, and enough technical freedom, let you focus on clean, secure, resilient architectures.
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.
> The clear specifications, reasonable timelines, and enough technical freedom, let you focus on clean, secure, resilient architectures.
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.
Wow. That was hardly ever the case in my experience in ~30 years in the industry. I enjoy(ed) what you’re describing too but… how common was this for you?
Different person but I’ve been a SWE about ten years now - my favorite work has been improving the CRUD stuff people have left behind. Like the time I took a proprietary document they were using to document their REST apis and wrote a converter to turn those into OpenAI spec compliant yaml files, taking a single tenant application and making it a secure multienant application so instead of 80 different app services we had two, coalescing all their data sources into one normalized DB. There’s a lot of work I’ve done I’m really proud of. I’ve always found myself being the “in the muck” guy fixing up and bettering legacy code, I’ve always felt that was really rewarding and interesting.
Thats badass! I like doing that kind of work, but I got very jaded about it over the years, because management generally acted as if the engineers took a shit on the conference table in the middle of a Zoom call with their most important customer whenever anybody attempted to address technical debt, as opposed to delivering more more more features. Unless it was billable agency work, in which case they’d hate it even more.
With massive amounts of effort and strong communication it was usually possible to get some tech debt time, but management acted like they were doing YOU a huge favor. It was kind of a career killer, generally.
Alternatively it was possible to get this kind of work done on my own time… but that’s a marriage killer instead of a career killer.
I burned out on trying after decades.
The fun thing is: I don't think this will change that much with AI in the mix ;)
Shitty devs without AI are slowly building BS that wants to be cleaned up.
Shitty devs with AI are much more fasterly building BS that wants to be cleaned up.
I have a similar experience as yours, just add another 10-ish years to the timeline. Literally my first job out of uni, I got to work on a legacy code base that was another 15 years older than that which had a ton of reports all built differently over time with slight variations in both code and end user output that really all wanted to be "the same" both on the code and end user side. So I slowly but surely, one user/PM requested change at a time would build out a common layer of abstractions and shared code that both made the code shared and easier to maintain (great for me!) and consistent for users (great for the PM/users).
Heck, I built a rudimentary TransactSQL parser that'd output stored procedure call graphs to graphviz and then I printed it and hung it on the wall to get through all this and having an LLM to just make a mermaid out of this in 5-ish minutes would definitively have been faster than I was ;)
With an LLM all of that would just have been faster to do but still worthwhile and useful on both levels. I.e. both code abstraction/reuse and PM/user level. Similarly there was a report that one of the devs had to actually run manually and it took him a full week, so users that wanted the report would only request it once a year-ish. When I was asked to learn how to run it from him I was appalled and just took that week to put it into a script that we ran and emailed the report out once a month automatically. They ran that report manually like that for like 10 years prior to that :mindblown:
None of this is an "AI thing". It's a good dev vs. not so good dev thing. And none of it really has to do with coding as in "writing source code" as much as "solving problems for humans" somehow and the "language" you do that in changes over time. If you're the guy that's tied up with the particular language: yeah LLMs are gonna eat your lunch. If you're the problem solver type that uses whatever tools they have available, you're still in demand.
Sure. If you get pleasure in fixing really badly broken shite, there’s tons of opportunity for that, but no money in it. Only money in new features, which is why the codebase is what it is. I have, on rare occasion, been the fixit guy called in temporarily to fix another teams work or get it out the door: that was fun. Short lived , but fun.
It's a shame, because there is money in it. It's just (like most sensible preventative maintenance) that the financial benefits are typically delayed. And when it comes to software, they're notoriously impossible to quantify.
Generally the only way to get it done in a corpo environment (and receive any kind of thanks or glory, as opposed to just torpedoing your own career) is to let things get so bad that production stability suffers and/or the team's ability to ship features is severely impacted.
...I realize I'm basically just echoing what you said, except more verbose. I'm venting!
About 80% of the time. It wasn't all about setting a new foundation, even making the existing one solid enough so you can have time to refactor the cruft was enjoyable. I had more trouble in my youth, as I didn't understand why the business didn't change to accommodate my 'masterpieces', and I'd like to think I started enjoying it more as I became wiser.
Disclaimer: I had to take an opportunity at management in cybersecurity, and haven't coded in about five years. It helps a lot when talking to developers about risk mitigation, though.
I’ve enjoyed it too.
One of the fun things to do would be to use the 8 hrs of work to spend about 4-5 hrs working on deliverables, which was usually sufficient to exceed expectations, using 2-3 of the remaining hours (I’d keep 1-2 hrs for myself), to make significant architectural improvements.
I’d bring in an improvement every 6 months or so and suddenly the app would get significantly better and I’d get credit for it.
> The clear specifications, reasonable timelines, and enough technical freedom, let you focus on clean, secure, resilient architectures.
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.
I was in that boat also
>I am one of those weirdos that found joy in business software. The clear specifications, reasonable timelines, and enough technical freedom
Where do you work! :)
I asked the same question, so that I could apply..
Sorry for the delay. Mostly midrange (AS/400) consulting at manufacturing plants, then full time jobs at a heavy machine dealership, and a bank. Changed to PC development after about 15 years of that, then management, and it hasn't been the same after that, ha ha.
I think the technical freedom plays a big part, not because you're chasing the shiniest, but because you're choosing the tools that will hopefully outlast you. One of my former employers was 'modernizing' their IBM i (former AS/400) systems by porting them to Java and Windows Servers, and they were having all kinds of problems. Couldn't even get the EBCDIC to Unicode conversion quite right. I was lucky not to be there anymore.
I smile whenever I see EBCDIC, probably because I never had to deal with it.
But I'm sorta old and some of my school's texts were outdated so they had programming examples and explanatory text that treated EBCDIC on equal footing with ASCII.
So unlike anybody who dealt with EBCDIC conversions in the wild, seeing that acronym actually takes me back to happier days!
> not 99% tweaking reports, trying to implement arcane and contradictory business logic, or fighting with shitty and frustrating externalities.
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
In a big production codebase with lots of active customers and data, architecture changes still feel pretty difficult even with LLMs accelerating them by an order of magnitude — migrating and backfilling data and all of that. That’s more complex than pointing Claude at a repo and saying “rewrite this in Rust.”
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.
in my experience it takes more than a day to sufficiently migrate bad architecture to a good one. like, many more days than it took to vibe the original implementation. having the prototype is good to understand what you want, but then going back and rebuilding it properly is a longer task. which is fine.
I think this is not "parallel universe". Selection bias: people who love to code often choose non trivial frontier work. They tend to grirf about agentic workflow the most. Myself included.
Popular HN fallacy: people who like to code are all exactly like me.
It sounds like you could be on my team, tbh.
I got into SWE because I like solving problems. I like solving problems using code. I also like solving problems using AI.
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
There's a lot of code I never wanted to be an expert on. I might care about an algorithm. I might care about the UX. There are other things I might let an AI tackle because I don't care about them.
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.
Devs were always feeding the machine to replace everyone else.
We were always replacing ourselves: Every "reusable" function or class or library you ever wrote represents lines of code won't get paid to write again. Every compiler takes jobs away from assembly coders.
The looming thing today is qualitatively different, because it's not formalizing something to reliably mass-produce.
> We were always replacing ourselves: Every "reusable" function or class or library you ever wrote represents lines of code won't get paid to write again
In theory, but not in practice. Instead, somehow we ended up writing essentially the same app over and and over, usually without reusing libs. You know because this time we'll use vue, or svelte, or phoenix, or htmx, or go, because this time we actually know what we're doing. so gotta start from the ground up and figure out how to store a row in a db as if we haven't done the same thing a million times before.
And that all under one roof. Now multiple that by thousands of roofs.
At least the devs were having fun and making good money writing all that code.
It’s a point that isn’t made often. AI is mostly saying the quiet part out loud.
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.
> 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".
And the housing costs are high because the rent extraction industry figured it out before the tech industry did
That's right-NIMBYism. California is expensive because of left-NIMBYism in the cities, and pastoralist old people in the suburbs.
(Also because of firefighters. And American urban planners who basically believe their job is to stop any project that anyone ever objects to.)
> Everyone in SWE is experiencing this.
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.
Yes. The people that didn't enjoy coding say that.
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 enjoy coding deeply, and my feelings are not that simple.
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.
As I like to tell others: before my job was 70% architecture & patterns and 30% coding & algorithms. Now my job is 100% architecture & patterns. While I still spend the majority of my time on architecture & patterns, and while I enjoy this aspect more than coding and algorithms, I am still working a "different job" than I was when my hands steered the code completely.
People say stuff like this all the time, but what does it actually mean? Like, isn’t the details of the code, at least at some level, part of the “architecture”? I’m so curious about what people who say this are actually making where they just sit around all day and dream up new lego structures without ever having to look at any of the bricks.
The bricks are pretty standardized? I mean, you're not reinventing singly-linked and doubly-linked lists from first principles no? When I program I think in terms of data structures, data hierarchy, and algorithms. I want to use a queue here, a lock here, a double linked list there, an array list over here, maybe a hash dictionary with linked-list chaining for collisions. It's only a few times in my career that I built a bespoke data structure from the ground up and they were all for very hot loops and one case involved working on a lock-free algorithm.
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.
I may have it wrong, but I think 'architecture' can include relatively low-level details: you could even design a program down to the level of what each function does, and still only be doing 'architecture', not coding, so long as you're not writing the code or specifying exactly how it should work.
> you could even design a program down to the level of what each function does, and still only be doing 'architecture', not coding, so long as you're not writing the code or specifying exactly how it should work.
So it all wishes and prayers? no wonder software quality is down.
> So it all wishes and prayers?
Software is built on wishes and prayers. Programmers don't ship pixels to be installed in monitors, nor do we (mostly) work with compiler internals. We write a detailed set of instructions and hope that a bunch of systems we don't necessarily comprehend follow them well enough to get the intended results. It works because some folks got together and wrote down their wishes and prayers so good that others were able to understand and follow them later.
We can be curious about how others approach problems or we can try to find the faults. I feel like this comment is a great example of jumping to the negativity rather than trying to understand a different perspective.
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.
> I ultimately think learning how other people structure programs is an enriching task
Learning from others is great. But there should be actually something to learn. When someone says
It means they’re not creating software. They’re just playing with randomly generating something.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.
What's the point of leaving a comment if you're not going to change your mind and they're not going to change theirs? Is it just to advocate for your position?
As you're being very vague, I write more about my opinion, which is based on a guess of what your approach. Basically the Q of P imply Q where P is the guess (and P is the thing you're being vague). I'd like if you can prove that either P is not valid or that there's a different Q that exists.
To further go with my point, it's like explaining how to make a cake and you said that instead of a recipe, it's fine to provide just the name of the ingredients and a general instruction like: mix everything together and put it in the oven.When we write code we sit atop a tower of abstractions. You're trusting your language's stdlib, its compilers, the syscalls offered by your kernel, the virtual memory implementation of the kernel. Your coworkers wrote code that you only interact with as a library or over an API of some sort. Let's not even think about the lossy abstraction of taking the ultimate bytecode generated and actually mapping it to the actual CPU pipeline that's running the code.
Modern computing is an exercise in trusting abstractions and then zooming into the abstractions and figuring out what went wrong when something breaks. Just because I wrote a web route doesn't mean I've audited the code behind the web framework, the event loop, the HTTP implementation, the TCP implementation and so forth. By your argument then I've already stopped thinking about my software and can't tell you if it's okay or not.
I'm sympathetic to both the perspective of sadness from the loss of hand coding and the perspective of joy from working at a higher level of abstraction, despite its costs in understanding. It actually seems very simple to me to hold both perspectives within me.
The thing you’re forgot to mention is that the trust in the layers below your code is not blind. Sure you treat it as axiomatic foundation. But when something breaks and you’re sure it’s not about your code, then you can investigate the actual part that is wrong in that foundation.
Once that’s done, there’s a matter of accountability where you can report the issue to the people responsible for that. Then because they know their own code they can repeat what you did and try to find where things went wrong.
That chain of accountability is the main issue there. We still have tower of abstraction, but no one knows about their layers. Accountability is lost and with it the trust is gone.
I never care about code, whether you wrote the majority of it or it’s just a glue that hold several library. What I care about is that the semantics of that module that I interact or build upon can be trusted, and when something wrong is an accident and not a daily event.
Even when you’re writing an application and not a library or a platform, you’re not free to write whatever. Your users depends on you, and themselves may have clients that depends on them. Being cavalier about what you write is just recklessness and in any other (real in some countries) type of engineering, it would be labelled as malpractice.
> That chain of accountability is the main issue there. We still have tower of abstraction, but no one knows about their layers. Accountability is lost and with it the trust is gone.
Right this is the main difference between our viewpoints. I don't work on completely open systems. A lot of my life I've worked on closed platforms, cloud runtimes, a mix of work that's been in my control and out of my control. I've filed my share of AWS tickets but ultimately most of them ended up not fixed by AWS. I've made peace with not having accountability for layers of the stack and am comfortable assuming that there are bugs on the foundation that I have to work around.
> Being cavalier about what you write is just recklessness and in any other (real in some countries) type of engineering, it would be labelled as malpractice.
An old critique of software engineers that I've always found flimsy and find that Hillel Wayne addresses well: https://hillelwayne.com/post/are-we-really-engineers/
I just design the fun part and let LLM deal with the chores.
For example recently I did tiny little hobby project, taking metrics of ESPHome/homeassistant MQTT and pushing it into prometheus. Coded the main parts, told LLM to do the boring part of setting up test suite, guided it thru making tests, it worked fine, tested it, then told LLM to do another boring task, taking all the datatypes it support and unit conversions (which is hundreds).
I got to do all the fun design stuff, but had code review assistant that also did all the boring tasks.
It's like making animation but you only have to make keyframes and initial coloring, with AI filling the blanks, just cutting out the tedium
Username checks out and thank you for writing out my own feelings better than I ever could have, even with AI.
There are many of us who enjoyed coding, and who enjoy the new paradigm, too.
Yes, sadly there are many people who are not honest with themselves because of peer pressure, FOMO and livelihood at stake.
What condescending self-righteous delusion.
People are really not as black and white as you are making them out to be. Most coders cared both about writing the code and the end result.
> and if you're not careful you will only learn how to prompt, not how the stuff works.
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.
That's a very simplistic view. I'm very good at writing tests and I'm happy to offload writing tests to LLMs, as long as they follow the structure I've provided.
You're saying that I must not be good in writing tests, irregardless on how I feel about test writing.
People who hated coding, technology, problem solving and only were in it for the money love managing a bunch of agents. Unfortunately, it seems like a lot of them were in technology. I hope agents quickly become good enough to completely commoditize any supposed function they perform, and forces them to leave this industry.
This is an extremely “all or nothing” take. I love computers and started programming in elementary school. I am having more fun now building software than I can remember.
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’m glad you mentioned flow state because that’s exactly the high I’m trying to chase with AI and why I like it despite programming since elementary school as well.
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.
AI Astrologists have proclaimed the month of the copium neverbeenbetterium.
You can love technology and still have thought that coding was a required path to get to technology and not the ends unto itself.
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.
It’s not the act of coding, you have to really hate problem solving, to love managing agents. Such people are only in it for the money, it may sound disparaging but it’s just true.
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.
There are infinite problems and agents help us solve problems faster. Sure they're not coding puzzles, they're actual problems actual human beings have, but they're still problems. Id way rather help a school district track its roster properly than argue with eslint. I'd rather get a rocket to the moon than triage garbage collection quirks too.
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.
Agent solve problems, you provide context
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..?
because writing the code was _never_ the hard part of solving problems.
Lol what? It definitely was. Gathering requirements is not hard, it is just boring and time consuming
If it wasn't hardest part, why pay so much to swe?
If agent did all the technical work of sending a rocket to the moon, and all that was left for me was deciding whether a rocket to the moon should be built or one to mars, which is better for our “customers” and I actually enjoyed that then I would be honest and say I was never into technology, which is completely fine, majority of humanity isn’t into it.
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.
> 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.
This is a false comparison. Agents cannot solve all the technical problems, so the entire statement falls apart.
What techical problems agent can't solve exactly?
agents are very very far from being able to send rockets to the moon just because you prompt them to.
Team of software engineers also can't send rocket to the moon. I don't get your point
I obviously implied code related problems. I know agent can't change a diaper or born a child, lol
You’re overlooking that the point of technology- and technologists- is to ultimately serve customers. The pleasure is incidental.
The point of technology is to make people’s lives better.
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?
What you and I are saying is synonymous. I’m simply reusing language that the post I’m responding to used. And also de-glamorizing/de-romanticizing things a bit, sure.
And given that fast food is the pinnacle of sustenance as a product and Hollywood has been content-brained since 1977, sure!
I’ve coded for 30 years. But I’ve always been motivated by outcome, imagination, intuition, & speculation. Coding was always a means to an end not the end itself. This is the funnest I’ve ever had creating software.
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.
That's reductive.
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...
> People who hated coding, technology, problem solving and only were in it for the money love managing a bunch of agents
In other words, people who just wanted to become managers.
I got into software later in life because I wanted to be able to send my kids to college, have healthcare and not destroy my body by 50. Up until my 40s I was a landscaper who had been an amateur programmer and technologist since my teens. I worked really, really hard to get where I am. I went to a bootcamp 5 nights a week for a year after breaking my back all day at my “real job”, while also taking care of a wife and two kids.
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.
It says a lot that you felt the need to recite an emotional sob story to try and attack my comment.
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.
Your response here says a lot about who you are as a person. Maybe it's time to step away from the keyboard and do some self-reflection and growing up.
I wasn't "attacking your comment", just trying to get you to relate to me as a human being. But I should've known that would be a futile exercise.
I'm good at my job, I like what I do and I work hard six years into my career. I just don't go home at night and do it as a hobby too. I know there are people like you would like to see me kicked out of the clubhouse for having needs and pursuing a career in one of the few middle class jobs left that pays a living wage. Sorry I'm not cool enough for you.
You should have just gotten an MBA, sorry for thinking that technology should be more than just earning some money for yourself.
https://geohot.github.io/blog/jekyll/update/2025/09/13/get-o...
Why did you expand it from what he said? I love technology, I enjoy being able to solve problems with technology and coding. I've always felt neutral about the actual act of coding, it was always just a means to the end and now I love that I can do more problem solving in less time and with less toil (from my view). But I guess I was just in it for the money.
> People who hated coding, technology, problem solving and only were in it for the money love managing a bunch of agents.
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.
FWIW I hate huge chunks of the industry + its inhabitants and am somewhat happy that my GPU can replace a significant amount of interactions with it and them.
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 love all those things, started coding as a 10 year old(30 years ago).
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.
People hate generalizations
Generally, people hate generalizations
Some people sometimes hate some generalizations.
> 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 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.
Most of the code I was paid to write is a burden. I've spent far more hours writing generic CRUD apps, debugging overly complex javascript tooling (almost always out of date) and other tedious grunt work.
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.
Fair. I think a lot depends on what type of coding you were/are being paid to do.
For me typing out code was always the annoying part. The fun comes from fixing problems, optimizing for speed, making the UX flow better and visuals look nice, all things that are on the level of software not code.
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.
> For me typing out code was always the annoying part. The fun comes from fixing problems, optimizing for speed, making the UX flow better and visuals look nice, all things that are on the level of software not code.
This always read like
Then you start wondering about the actual level of education of the person and their claims about writing a nice novel.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?
sure, it works like that if you're still the one doing the thinking and figuring out how to do things, and just using agents for the boring templating and to translate your pseudocode into syntax-free real code, without you having to look up the arguments for every call that you haven't used recently, etc.
but most SWEs are being pushed to delegate the _thinking_ to agents.
so instead of "here's a solution to load this page faster that I came up with: X. Implement this in JS", which is along the lines of what you describe, most people are doing "this page needs to load faster, what's the best way to do that?" claude: "X". "OK, implement X." What fun is that? What's the point of developing skills and acquiring knowledge if you can't use them anymore. that's the loss.
If all you care about is having pages that load fast, then yeah, agents are great. If you care about being the architect who figured out how to load pages faster, then agents take the wind out of your sails.
Your example is great because I literally just did this:
In the past I would have cracked open a profiler, and spent a couple days probably tuning this thing. I probably would have gotten a big dopamine hit and felt really accomplished when I finally got buttery smooth performance.But I'm happy to delegate this task to the agent, because this is one step in 1000 steps I have to accomplish, and no I can't just tell the agent to accomplish them all from the start because I don't know where it's going. I'm very very happy I don't have to waste a day making the graph load faster, because I can get back to my real work, which is not making graphs load faster.
At the end of the day, solving that problem will make someone's use of my project a little better, but it's not adding value to anyone's life -- that's what the project is (supposed to be) for. In the past, it was those little problems and 1000s more like them that would derail a large project. Today, they are mere bumps in the road, so the large project has a better chance to come to fruition.
is this a personal project, or your day job?
in this context, those are really 2 different things. what people are lamenting is the fact that "I built a career on building these skills because it's something I care about doing with my life, and now I'm being relegated to QA at best, or replaced at worst". That's completely different than, "I can now get my personal project out faster, or maybe create 10 personal projects, and can have the agent take care of the parts of the project I don't really care about doing (like improving graph performance) and focus instead on the overall design of the project, more features, etc.".
It's like going through med school and residency to become a doctor, and instead of practicing medicine, you're feeding prompts into an AI and passing on the responses. Sure, on you might appreciate not having to field medical questions from all your friends and relatives ("personal project"), but that's not your career.
My day job, but I am not a software developer (I'm a teacher), although I was in the past.
I can appreciate that perspective, but at the same time: it's a job, the whole arrangement is you sign away your autonomy, you do what you're told, and people were paid more in one year than I've made in 5 for this. Now the balance has been disrupted, and that's a shame, but the gravy train can't go forever. Moreover, this particular job, software developer, has changed fundamentally again and again over the years, and those changes definitely disrupted people's careers, so although the blast radius is significant, this too shall pass.
And we're not exactly talking about doctor-level of training here, it was only about a dozen years ago that it was very fashionable to land a 6-figure salary at a top Silicon Valley tech firm after having only attended a bootcamp, for a job that was characterized as being a "code monkey". The whole thesis of the "Theil Fellowship" was that the job was doable without any formal education at all. At my institution, the CS major was the most popular because it was the shortest path to the largest salary. Now that equation has changed, no one wants to be a CS major, enrollment is through the floor.
> you're feeding prompts into an AI and passing on the responses.
Okay but is the job writing code or building software? A lot of people seem to think the job was writing code, and that's what they liked about it, but it was always the case that if you like your job it's not going to pay well. That software engineers were paid so well to do something they love is an aberration, not something to expect forever.
As a teacher, I definitely understand this. I love my job, doesn't pay well, and it's also changed fundamentally as well in the last year. So I empathize -- I used to grade student papers which was rewarding; then all of a sudden I was grading LLM output, which is not. A lot of my colleagues have complained endlessly about this, have made it their job to become the LLM "police", or have quit the profession entirely because like you said it's not how they started in this field. It's bleak.
But you can also adapt to the times. We just completed a module where students use AI to create games in Lua, and they made the best games I've seen in years, each one different and unique and fun. Before the class was "we'll spend 4 weeks and hopefully we'll get to animations if we're lucky". Now the final projects are all animated with interesting gameplay mechanics, graphics, animations, and they're all different despite them each using AI. I like this situation better.
Soon enough the clankers will automate that part too. I'm eager to see how "liberated" those people are going to feel then.
Totally, I can't quite imagine an AI automating my use of AI tough, how it would infer what prompt I would use at any given time, the problem space is infinite. Am I adding another *arr app to my stack, are we building a SPA, a home automation plugin, whatever. These are things that I want to do and AI helps with that.
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.
It is remarkable to me how few people seem to understand it is a much smaller step for AI-generated software to kill the need for companies of idea men than it was for them to kill the need for people writing code.
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.
> 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.
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.
CEO will just be 'human scapegoat' to catch flak when the AI messes up. You'll get a nice comp as long as it doesn't.
I don't think that's ever going to happen when models still hallucinate and prompt-refuse, especially since one of the two SOTA model makers insists on aligning their models toward a progressive ideology.
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.
> 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.
Because you are singularly right about the issue? Have some intellectual modesty, brother.
I think it's fair to expect tha AI capabilities will more likely end up far over or under that magic threshold than close to it. Given the way that technology develops, it's probably going to be over.
Its nice to know others feel this way. I do feel some kind of despair when I think about the kind of work I used to do even just a year ago compared to now. A coworker of mine took FMLA recently to figure out what they want to do next because they no longer feel even the slightest fulfillment from swe.
Interesting take and I certainly agree that lots of people think that way.
In my case I had prior experience with tech support and I saw lots of people struggling to use computers and it's interfaces.
So my joy comes equally from coding or seeing people relieved that their problems were solved.
With that mindset that I'm able to fly with Claude or codex.
And even doing my comp sci more than a decade ago many professors told that I should worry more about the outcomes and what problems I'm solving, instead of building this life game around coding only. So I'm grateful for them.
Maybe that can help other people getting out of this dread feeling whenever they are using A?
My frustrations changed, nonetheless, with AI. When it forgets cases that the LLM session and I forgot to cover, that for me is pretty obvious if I had or gave more time to evaluate and review things.
This week I realized that running a separate session and asking dumb questions about the plans and recent pull requests helps grounding again.
I have to agree, but I think some people find the joy in finding their solutions in code rather than from code…meaning that for some it is the puzzle solving that is their interest. That can be rather hard to find an analog for, especially if coding is the skill they are good at.
The lack of AI mental model between conversations or even within the conversation is a definite frustration. “No, I didn’t just decide to entirely change everything I’ve planned based on one ambiguous word…”
Your use of a separate session for q&a is actually something I find interesting in the new Granite Switch architecture’s KV Cache Normalization, where the kv cache is essentially frozen while the model performs certain other tasks in order to prevent the mental ‘conversation’ from getting polluted. However, I’m also assuming what you mean by “helps grounding” is that it helps keep the agent grounded in the context by not introducing a bunch of external information, which is certainly beneficial.
The other side is management, instead of being team leader for an offshoring team, which at least were human, those that manage to keep their jobs will be a robot manager.
I certainly did not come to computing to be a manager, in fact I already had that role and freely downgraded myself.
Of course this does not undo any grief, dismiss it, or better their situation:
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.
That just makes it feel like I wasted a lot of time. I spent years eschewing social gatherings, foregoing other opportunities, and depriving myself of things I enjoy, to cultivate a set of skills. Now I feel like they are rapidly becoming largely irrelevant (vuln hunting, kind of on the leading edge of ai displacement, not sure bug bounties as we currently know them will exist in five years time). What's more I don't feel like I'm positioned to remain employed in the field. I didn't expect my status or stability to be thrown into jeopardy so rapidly. I wish I were as selfless as you. I'm not sure many people are, maybe the world would be a kinder place if they were
that probably makes it feel even worse for most people
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.
> It's easy to say that such people are prideful, arrogant, only cares about being better than others
The most vocal denouncers of egoism are utterly pathologic egoists themselves.
All this supposed fun, is always something fairly simple that would take maybe a hundred or thousand hours of practice to build, it was never out of reach of them, they just felt they didn’t have the time or motivation to build that skill. (even a 1000 hour project can be done in 6 months outside of work time) Now they don’t need it. If there is any fun in this, it is an extremely shallow, garden variety sort of fun that comes from consuming content. They will never experience the deep fun that comes from being immersed creatively in a project, which I would trade anytime for something shallow.
I don't think your expectation is reasonable for the majority.
The majority likes watching slop all day and rot to death. Their tyranny is not something that should be celebrated like your original comment.
Uh...
Edit: I'm at a loss for words.
Tell us more about your mastery of every creative domain.
This comment is the "Eat your broccoli because there are starving kids in China" of SWE.
No, that's not quite the same.
To the person who disagrees that it's not the same: I'm not sure how to prove a negative.
We aren't suddenly seeing 10x the software output that these 10 joyful people are producing. I think it's more likely that 10x people are suddenly trying to figure out a way to low-effort scam money with AI. And a few of those people have teamed up with VC's to make AI products to scrape the bottom of the pond for desperate suckers looking to scam money. There's 10x the headaches for existing projects that suddenly have to deal with low-effort bugreps from people looking to show that they are now "expert software engineers" because they can write "go find the bugs in Project X" and an AI will possibly do that.
There are some interesting things being built by AI, but 10x the number of joyful software engineers ain't it.
Not everyone is going to publish their creations, not everyone is on HN, and not everyone on HN is going to shout from the rooftops about how they are in that larger group of newly-enabled people.
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.
But are we actually seeing those results or are people just burning 10x more money?
What I'm not seeing is news promoting how we're suddenly doing 10x better in any aspect of our daily lives, nor 6x or 4x or even 2x better. You'd think if we were suddenly getting 10x more productive, even if it was never published in the news, we'd see something, anything that affected our lives in a positive manner.
We're not.
Or at least, I'm not. And a lot of other people seem to be in the same boat.
I think you're 'putting words in my mouth'. I didn't say that people were becoming 10x better or more productive, just that there are 10x more people being enabled by AI versus programmers who see AI as destroying their jobs.
[citation needed]
Doubly so when you consider the history of every previous advance that was supposed to free people from the tyranny of needing a developer to make software, from COBOL forward. Some people indeed made their own applications, while everyone else rejected doing that and preferred to pay someone to do it for them.
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.
> 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.
We had our work stolen and they're squeezing us with it, all for garbage, what in the world are you talking about.
> And also that on the other side of that grief is something new and potentially exciting
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.
> 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.
Would you look at a piece of code elegantly using high-level abstractions to model some problem written by a human and go "see, look, code is beautiful"... and then look at the exact same code written by an LLM and say "this is slop"?
> Moreover, it's downright offensive for a purported expert to defer their expertise to a SAAS black box.
With programming, it's just the opposite, though.
Programmers work with abstractions. Understanding the full system down to how atoms work is too much to think of when e.g. writing an API handler. So instead you work with modules and their interfaces.
Pre-LLM, if you need some task done, you wouldn't write it all out yourself if there was some external package which did the work for you.
Of course, when using an LLM, you don't need to surrender your ability to think about the problem and solution.
What is an abstraction? An abstraction is far more than a black box. From a previous comment:
> High-level programming languages use well-defined, limited syntax to encode program behavior in a deterministic way. Compilers translate these directions into programs that should have identical behavior between compiler runs and between different compilers. (If they don't, that's a compiler bug, or else an artifact of UB.) This means you can structurally and provably rely on the high-level programming layer without worrying about what's underneath: the assembly is abstracted away. You also get the massive benefit of being able to tweak and improve the assembly layer independently of the programming language layer, so long as the contract between language and compiler is not violated.
> If AI is a "layer of abstraction," so is working with a bunch of contractors to make your app. The definition is broadened to the point of absurdity. (Solely: "I don't have to think about code anymore," which only touches on the most superficial aspect of an abstraction.)
With a good high-level abstraction, I can mostly think on the level of the abstraction, but I can also dive down through the layers and figure out what's actually going on in each one.
> Would you look at a piece of code elegantly using high-level abstractions to model some problem written by a human and go "see, look, code is beautiful"... and then look at the exact same code written by an LLM and say "this is slop"?
Using is one thing; inventing is another. I have seen no evidence that LLMs have the capacity to discover and leverage new abstractions. To me, that's the most exciting part of programming: finding ways to reify powerful concepts and then turn complex tasks into simple ones, ad infinitum.
The Luddites literally turned to violence when their profession was replaced by better technology. This has been happening for generations to every profession. There is joy in learning new things and developing new skills though.
The Luddites did not resort to violence (unless you count sabotage of machinery as violence), and they were not protesting against better technology. They were protesting against the exploitation of the workers that these machines enabled. Formerly, weaving textiles was a relatively well paid and highly skilled profession, then these machines largely automated the process. This lead to the means of production being owned by the wealthy, employing badly paid and overworked people, to make themselves even richer.
Yeah its funny how not only history books but even language itself is changed to suit the victors of past conflicts.
The “violence” was a simple act of sabotage, nobody got hurt except some peace of machinery and the wallets of the rich exploiters.
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.
It's so funny watching tech workers spout propaganda about the Luddites.
Brother, they're you. You just don't realize it yet.
Not everyone. I’ve been a solo entrepreneur since 2015. Developing stuff now feels like I can unwrap a present every day.
I never cared about the social aspects of it other than releasing the odd plugin I developed.
My social life is entirely offline.
> Everyone in SWE (who got into it for the love of writing code) is experiencing this.
Speak for yourself.
> Everyone in SWE (who got into it for the love of writing code) is experiencing this.
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.
Yeah the AI companies slurped up all our work and are now selling it back to us in the form of trained LLMs.
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.
When I forecasted open-source is going down because nobody wants to expose himself only to be robbed off credit and authorship, he did not initially understand what the claim was. Two years now.
Indeed. This whole LLM has been a planetary scale theft of labor. They stole our labor and are now leasing it back to us as rent-seeking on the means of production.
Those of us who refuse to use that stolen labor are loosing our jobs, fast.
It's not even limited to software engineering.
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..
“If you want a simple model for predicting the unemployment rate in the United States over the next few years, here it is: It will be what [Alan] Greenspan wants it to be, plus or minus a random error reflecting the fact that he is not quite God.”
-- Paul Krugman, 1997
Note this does not say "it will decrease if we suddenly invent new productivity technologies".
Jokes on them. You don’t need MBAs any more. You just need the cash or a credit line.
If you mean publishing open source code, they didn’t steal it. We gave it away. Generosity has always been vulnerable to “exploitation,” but some of apparently understood that risk better than others.
My jumping off point was OSS, but I was actually talking about the entirety of human knowledge being used to make a few people insanely rich while sending everyone else who made the LLM training possible to the bread line.
No, I reject this framing. Copyright violations are orthogonal to labor theft. Labor theft can include copyright violation, but it doesn’t have to. My MIT licensed code may not have been violated (it probably was though; where is the copy of the notice that my license requires?) but my labor was for sure stolen.
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.
You already valued your own labor at zero by giving it away for free. How is taking something you give away for free suddenly considered "stealing"?
The same way that Fitzcarraldo stole the labor of the Amazon Indians when they moved the ship over the mountain in order to profit from the rubber.
Previously when I licensed my code with a liberal license I got something back, credit, contributions, connections, job opportunities, etc. The AI companies stole my labor, then they demanded I pay them, and finally they took my job.
People don't "value their own labor", sociopaths do.
Stealing labor itself isn’t actually a thing. Your wages could be stolen, except - what wages? Maybe the output from your labor could be stolen, but not if nothing material was lost. If you can’t point to a specific intellectual property violation, you have no case.
Did they steal Wikipedia? Wikipedia is still there.
Stealing my labor means that I do all the work, and they take all the profit.
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.
It sounds like you're making a claim to someone else's profits with a very thin rationale for why you had anything to do with it. Basically, claiming something that never was yours and has next to nothing to do with anything you worked on.
My labor was mine. Even the worst capitalist ideologues admit that workers own their own labor and the sell it to their employer for wages.
You are basically arguing in feudalism. My labor belonged to my landlord and they had full right to take all the profit for them selves.
That was the point of FOSS and copyleft: the abolition of copyright on code allowing anyone to take code from any place to use however they please.
The entire concept of copyleft and free software depended on copyright: it isn't "take my code and do what you will", it is "here is a very explicit contract that demands you credit my work and even requires you to contribute back in return for my efforts". And like, maybe a world with no copyright at all is better! But that isn't the world we are in: it is no more legal for you to have an AI read my code and then port it into your codebase (which could involve training as an intermediate step or could merely involve the AI searching for it and loading it into its context) than would be for you to directly read my code and port it into your codebase or even for you to utilize a deterministic tool to read my code and port it into your codebase; you only get to do those things if you retain my license on the result and abide by its terms.
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 depended on copyright as the legal structure to uphold itself, a means to an end, not the end itself. As another comment in this thread nearby says, the point was to support the four fundamental freedoms and ultimately a world without copyright on code at all.
And on the first part about AI reading your code or training from it, it does seem like it's legal now, while you as a human reading and porting is not.
Yeah, I never contributed to FOSS to get credit; I contributed because I wanted something improved which I thought was not very good and thought that sharing the improvements with the world would be good (sharing is caring, after all). From a Free Software perspective, I actually quite like that Generative AI allows people to create the software they want and adapt it to their wants and needs. Free Software is about freedom, after all (free as in free speech). Unfortunate GAI tends to be quite resource intensive, which is bad; but local models are getting better all the time, so hopefully eventually anyone can run a GAI model on their own computer and have it create any and all software they want.
It's a bit like Bob Ross' Joy of Painting: if everyone can paint then anyone can make any painting they want to see.
Copyleft comes with restrictions on code - you have to keep it public. It's the BSD licenses that basically don't.
The end goal is not adoption of FOSS licenses, the end goal was the Four Freedoms ( https://en.wikipedia.org/wiki/The_Free_Software_Definition#T... )
- The freedom to run the program as you wish, for any purpose (freedom 0).
- The freedom to study how the program works, and change it to make it do what you wish (freedom 1). Access to the source code is a precondition for this.
- The freedom to redistribute copies so you can help others (freedom 2).
- The freedom to distribute copies of your modified versions to others (freedom 3). By doing this you can give the whole community a chance to benefit from your changes. Access to the source code is a precondition for this.
Note that "preserve licenses" and "preserve attribution" are very much not in that list. Well, the Four Freedoms are what these AI doohickeys enable by being able to in theory reverse engineer or replicate any code, making all code effectively "public", and to bypass any copyright issues around distribution. The fact that AI blowing away the existence of proprietary code also blows away the licenses of FOSS licensed code is a harmless side effect of getting to the Four Freedoms, which was the Free Software goal all along.
If any of this is surprising, then you didn't understand what you signed up for. The AI monkey's paw has curled and you've got what you asked for, sort of.
And none of those freedoms are guaranteed today for all software but we effectively killed the the mechanism that guaranteed them at least for open source software.
it feels like a transition from artisan level work to factory based industrial manufacturing.
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
And the same thing will happen to devs that happened to factory workers. Jobs will get moved offshore, and salaries will drop, mass layoffs. And SWE's don't have unions like factory workers did.
Maybe it's time for SWE's to unionize.
It's long since past, and it's not like unions would've saved the industry from AI, like Europe has unions too (and their software industry is quite poor compared to the US in general, partially due to higher regulation on companies and workers both).
It doesn't have to be one or the other. Having a machine that does most of the coding for you means you can have more of the collaborative, human input. You can focus on the UX, you can fix all bugs, you can experiment with new ideas faster than ever. There will be zero excuses for poorly done work - think of all the half-assed corporate apps you have had to use during your career. I think this is the start of a great era in software development.
This is the end of software development and the beginning of software management. The problem is you don't need a lot of people to do the later. There are going to be a lot of devs out there vying for the same jobs in the future drastically driving down salaries.
I am inclined to think this way in a commercial setting, but for open source contributors, where their work was done for the love of coding, I can understand feeling like the end has arrived. In my company, I'm building the mechanics of factory-coding and trying to get the team think in higher-level terms: what needs to be built, what does it need to do when it deploys, and how can we build the factory that gets that done repeatedly? To me, building the factory is an interesting problem and agentic coding does not eliminate the higher-level design that absolutely must be done, all of which is, to me, the interesting part of software dev.
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.
And when before iterating on a stupid web page layout might've taken weeks of back and forth - now it can be done during a single meeting.
Or during a sales meeting, an experienced engineer can have a quick prototype done to show the client.
I imagine it was what it felt like for miners when they went from working next to each other with pickaxes, to operating massive machines by themselves, although that timeline was much longer and less jarring, taking a generation or two.
Except no miners went down the mine after their day job to work for free and for fun
The miners also weren't used to train the machinery replacing them.
You write sixteen lines, and what do you get?
Another PR and technical debt.
Eh, maybe the geologists did.
I wonder if we will see new platforms emerge around human-only collaboration. There's nothing stopping people from continuing to work on projects in the traditional way. It's just harder on big platforms like GitHub which are designed around public bug reports and pull requests.
I hope so, and not only for code but art and even mindless entertainment. But you really need to draw a very hard line or there will be a line of useful idiots insisting that their "minimal" clanker use is OK and that they should still be accepted in your community. And then the next one will want just a little bit more allowed and before you know it your community is no different from any other on the net.
One of my favorite hobbies used to be to explore GitHub and see what interesting projects were created or updated. I loved to checkout and learn from others.
Everything is weird now.
There is this belief amongst software engineers that humans are always part of the process and that AI prompting is just the new IDE. Yet, every new model release is even more optimized to replace programmers and is pushing us one more step closer to that reality.
Right now (Fall 2026), we focus on empowering agents and LLMs and that's the current challenge facing software. However, once we solve that, the LLM really can kind of take it from there.
I know I'm being dramatic when I say this, but it really is how I'm starting to become jaded from all of this AI tech. It looks more and more like we're becoming a world full of really brilliant humans who are bored because there aren't any intellectual problems left to solve. That is not a utopia. That seems like a dystopian nightmare.
You have to wonder if this site will go next.
He's dead, Jim.
Just like with mathematicians, is it all really just about people losing a random, replaceable job and what they enjoyed about it?
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.
> One wonders whether a generation that demands instant satisfaction of all its needs and instant solution of the world's problems will produce anything of lasting value. Such a generation, even when equipped with the most modern technology, will be essentially primitive — it will stand in awe of nature, and submit to the tutelage of medicine men.
- Eric Hoffer
I think we'll be okay as long as we don't outsource our thinking to technology.
Wait... What's that you say?
> Just like with mathematicians, is it all really just about people losing a random, replaceable job and what they enjoyed about it?
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.
That's the way the world's been since the integrated circuit was first developed, for better or worse.
No, it's not.
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 point isn't some people not understanding whatever stuff with limited effects, it's no people understanding anything important.
Using an LLM doesn't mean "I must write this code in a way that I cannot understand it". I'd encourage using the LLM to write code in a way that you can understand.
> When corporations own it exclusively, they can do whatever they want with it.
This is an argument in favour of open-weight models.
but so many things we have are black boxes now. do people need to understand how cars work to drive them properly?
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).
The difference is that people who make the cars and the mechanics who repair them still understand how that black box works.
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?
probably not, but my main gripe would be that there would be no way to punish anyone if something went wrong with the car, not (necessarily) that human judgement or expertise is still needed as part of the car design process.
So yes, maybe we will all just exist to get sued at the end of the day.
A more apt analogy would be if self-driving cars fully take over, such that most people don't even know how to drive, let alone how an engine works.
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.
Yes we do need to understand. We are already seing how things being black boxes is used against us via enshittification.
> What happens when you don't understand them anymore?
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).
I used to love making the occasional event poster for friends and signs for local businesses as an aside to my day-to-day work. Early in my career I was in advertising and design, and still love my art.
But in the last year you just don't get these jobs anymore, due to AI. It's everywhere and I totally understand the feeling uncreative people get when they are finally given a tool that lets them create something.
What makes me depressed however, is seeing that the vast majority of people are perfectly fine with AI slop posters etc... Highlighting to me that they likely didn't actually value or appreciate the good design to begin with. A lot of people don't care about the human element to anything.
> A lot of people don't care about the human element to anything.
I think this is a crucial point. The AI era seems to bring out two distinct movements in people: Those that always saw the human as a necessary bottle-neck, and those that actively sought and enjoyed interacting with another human brain.
> Those that always saw the human as a necessary bottle-neck, and those that actively sought and enjoyed interacting with another human brain.
Yep. The problem I feel is that humanity and society evolved via human interaction over the span of tens of thousands of years. And then in the space of a decade some of the people who benefited from that development feel so comfortable in their lives that they seek to literally end that method of progression for our species.
Yeah the most depressing part of the AIpocalypse is the realization that most people simply don't have any taste whatsoever. At least that's the only way I can explain why these horrid slop posters are thriving instead of essentially killing whatever business is putting them on their windows. It's not even just the missing human element (but it is that as well) but the level of quality that the average person seems to be OK with is just so low. Lower than what pretty much everyone should be able to make by hand.
I never understood this. It's like home cooking, I don't stop just because a professional chef can make it better. Seems like many OSS contributors wanted to have their cake and eat it too, they're not really contributing for the love of the game but for external factors like credit or admiration or clout.
Would you keep cooking if a Star Trek style replicator existed that could more or less cook anything in a minute? What if all of your friends and family were satisfied with the replicator food and weren’t willing to wait for you to cook anymore?
It's like ordering food from outside versus cooking food yourself. Yes, you could order it and it might be appreciated but cooking something for others will, if they're good people, have them definitely appreciate your effort. And there's a difference between cooking for sustenance daily, in which case I too would use the replicator just to save myself the effort, and cooking for fun and special occasions.
This is analogous to coding as a profession versus for fun. For work I'd use all the tools at my disposal to get the job done because that's what matters in a corporate environment. And for home, well, I'll code some things and vibe code others, depending on how I feel about a certain project. I think OSS maintainers lean too hard in setting themselves and others too high of expectations such that they burn out like this, many such cases, when in reality they shouldn't take it so seriously. Work on your code when you want to and leave it alone when you don't, the license literally says "as-is," so follow that philosophy.
i find much joy in building things that would have cost me 5 figure USD in dev pay or months and months of work. there are two sides
How is that being disillusioned?
I am always doing everything alone, so I am liking the LLMs a lot. My own code is very good, but nobody but myself cared about that. Even the one job that I got which allowed me to gather the capital needed to move into trading feels like complete luck. This is a new golden age for entrepreneurs and apocalypse for wage workers. Nobody whose status depends on his own productivity would complain about it going up sharply.
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.
"This is a new golden age for entrepreneurs and apocalypse for wage workers"
"right now the future does seem bright"
This reads like "Fuck you, I've got mine." Who cares about those wage workers anyway, right?
I think you’re heavily miscalibrated. Anyone can use LLMs, the floor has risen drastically. What supposed productivity you achieve, every one else can because there’s 0 skill to using LLMs, this is probably a much darker future ahead for you than the rest, you’ll have to keep constantly agent maxing to try and outcompete millions of other humans doing the same.
I'm not sure this is credible. Using AI very much seems to be a skill with its own craft and best-practices. I could imagine a similar argument was made by painters about photographers. Yet, we believe good photography takes skill.
Analogy time!
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.
People are afraid to do work on cars (and mains wiring, and plumbing) because even though the work is relatively simple the consequences for fucking up are so severe (severe damage, injury, death, etc). That's most of the reason they pay someone else to do it.
For the vast majority of software that isn't the case at all.
Most people don't understand computers and do not even know what is dangerous and what is not. They won't suddenly become vibe-programmers lest "AI deletes all of the personal photos and will send them somewhere".
!RemindMe 6 months
No need, I still have vibe coded stuff from the times before monthly plans for Claude and Codex over a year ago. I can jump into most of them and just start adding stuff without any major issues - they have better test coverage and CI than many projects at my day job :)
It's surprisingly easy to do that shit properly when budgets and time aren't an issue.
> agent maxing
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).
LLMs are constantly driving off the cliff. Constantly. They require some actual agency steering the ship.
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.
lol wat? there is a very clear gap in output between what a good swe can produce with ai versus a bad one, let alone a random person who doesn't even know what javascript is.
One thing that strikes me about what is going on: we have been promised this renaissance in software, yet, it been 2-3y of accelerated agentic/LLM development and I haven't see anything great coming out of the LLM developed software world that is worth looking at. Let's take open source as an example: No new 3d/2d CAD packages that are more capable than Blender/FreeCad/KiCad, nor Blender/FreeCad all of the sudden became much more capable than they were just few years ago. I also want to point out,that I haven't seen any new LLMed software in Electronic/ElectroDynamic simulation, something I would LLM to excel at, beyond what already existed before. No new software that is equal to Adobe Creative Suite (that Vibed coded replacement is trash IMO), No new Operating Systems that are taking over. No new utilities that are super useful in day work. We are using the same software, we've used before LLM arrived.
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.
One of the things I've started noticing at work is that for the last few months I was banging away at having multiple streams at a time. But over time, I've reached the point where there is some human blocker on the stream, be it QA to test it, other humans to have attention to pick it up for the next step, deployment of some sort, or in some happy cases, it made it through to deployment and the problem is now solved. In the past couple of weeks my multistreaming is toning down and more often I'm back to one stream of work. Often a much higher level stream than I would have been working at a couple of years ago, and I'm moving faster in a number of ways. Some of that "faster" has been burned up in things like changing environments that I previously just wouldn't have done, though.
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.
I think 2x or 3x is maybe too generous. Obviously pure code output is extremely high but I would think for something to truly be a 2x a company would either have to be able to effectively get rid of half its SWE workforce or produce enough to make some notable increase in revenue.
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.
We might be producing shovelfuls of code at 10-20X the previous rate, but we're not producing value proportionally, maybe we are even inversely proportionally.
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.
People say this but I mean, don't you still want that 5 to 15 percent of time back? I certainly appreciate my time back with my LLM work, but then again I work remotely so there are lots of benefits to extra time around the house while the LLM churns away.
> don't you still want that 5 to 15 percent of time back?
Guess it depends on the downsides.
Also, it's not like the time spent with an llm is zero, so maybe you save 2-7 % (depends on the task, of course)
It depends on the role too. I had jobs where I was in meetings all day and others where it was all asynchronous and had essentially no meetings, and so in those latter cases I could get a lot of time back.
> Let's take open source as an example: No new 3d/2d CAD packages that are more capable than Blender/FreeCad/KiCad, nor Blender/FreeCad all of the sudden became much more capable than they were just few years ago.
I know nothing of Blender but in the FreeCAD community there was a little disagreement about use of AI and it has taken some time to find a policy most people don't mind (one of the original authors stepped back from maintaining because he disagrees with this policy). Plus they have also traditionally had quite a long release cycle.
Nevertheless, FreeCAD is much more capable than it was just few years ago (largely through slogged human effort in the main) and the velocity of change is very evidently increasing with AI.
However, CAD applications are quite a difficult example to pick, because bRep kernels are hard, FreeCAD is not in control of theirs (it's OpenCascade, also used by PrusaSlicer/Orca Slicer for STEP conversion), etc.
AFAIK OpenCascade's development has picked up recently; whether they are using AI I don't know, but I expect they are not accepting AI commits from outside sources.
I have. I use a MacBook and now lots of games that didn't have native ports now do, via vibe coding and AI reverse engineering. This would have taken years before and now anyone can ask their LLM to do the same for their favorite game or app to run on whatever device they want. You say the vibe coded replacement is trash but in some time it'll because better simply because users will iterate on it adding features that they want, not what a corporation thinks they should have.
OP of the top reply. As I have mentioned, RE process is pretty much automated thanks to AI. But once again, RE existing software/games and not the same as developing a new software/games. AI being great at RE is an inherent property of AI itself.
I think this comment is premature. It hasn't really been 2-3 years accelerated LLM development. Even 6 months ago, LLM development was much less refined than it is now. There's been a few years of figuring out how to use these tools (which are constantly changing) and maybe a year of using them effectively.
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?
>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.
My projects are so niche that almost nobody would ever care... but think that might be the point. Maybe there aren't any decent MS Office or Photoshop clones but there's also a bunch of really interesting niche software being created nobody else would ever care about.
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.
I worked for mid size Insure tech company that has good footprint in the industry and two years ago, LLM were good enough for us to refactor & rewrite major tech debt. Even two years ago, models were already impressive & capable. For me, the timeline been two years and counting.
A year ago I was also doing major refactoring and it saved several months of time (probably wouldn't have even attempted it otherwise). But LLMs still weren't as capable as they are now for more interesting tasks than that. I struggled with an LLM assisted conversion project, eventually just shelved it, and started it again recently and the difference is quite noticeable.
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.
I thinks what's more happening is bespoke software made by people for themselves mostly, often with a user count of 1. That's quite a huge change, for users of that software. Of course such things are not replacing existing apps.
> buggy, as they were 4-5 years ago
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.
The quality of everything has gone downhill -- software is just part of that trend. Everything is about extracting as much value as possible for as little cost as possible. We've optimized everything, including software development, to that. Quality requires a level of care that society is no longer able to achieve.
I recently re-read Venus, Inc. and the number of times that the satire didn't hit because it read too close to our current reality was quite depressing.
People say that, but it doesn't hold up. We spend more money than ever before on software development. Just not on quality. There is something else going on here.
You'll need to backup the statement that we spend more money than ever on the same quantity of software but not getting more quality.
This is the great paradox that many people have observed: why are we still using frustrating, buggy, hard-to-use, expensive software all day?
Because LLMs aren't good enough yet to qualify as enterprise support.
The last time I encountered a bug was in some open source software. I asked AI for help. The AI read the 350,000 line project on GitHub and said "yes, that is a bug, but nobody has reported it", and then showed me the code related to the bug.
I cloned it and fixed it myself.
Similarly, I've asked questions about open source projects with poor documentation. The AI just reads the entire source code and finds the answer.
It will be interesting to see if having access to the source code becomes more valuable because of this.
Yes, I think having access to the source code is increasing going to be a requirement for enterprise software... and I think that's going to come from security rather than bugs.
The problem with self bug-fix is that it leaves you on a unique fork. Repeat that 100x for 100 different users, and now how does upstream add a new feature reliably? Everyone's going to merge into their own fork?
My guess is we'll still be using upstream-authoritative, but at some point user-initiated patches will flood over the dam and become the new normal even in formerly closed-source enterprise.
Because source is going to be immaterial (decomp will be good enough) and the security benefits of customers being able to self-scan are pretty high, especially in an age where time-to-exploit approaches zero.
... or it'll just hasten the move to SaaS.
Well, I made a PR and it was accepted.
Yes! It's just repeatedly asserted that vibecoding is taking over, going to replace us all. I have yet to see evidence of it (no, marketing stunts aren't evidence).
We have seen a huge proliferation of vibed audio plugins. Most of them are very bad or replicate things DAWs can already do. And vibed video games. If anything it has become harder for many audiences to find good software because algorithmic curation is driven by gameable engagement metrics not quality.
Hmm, that may well be true but I don't model with any of those packages. I model with `build123d` or `openscad` using Claude Code to do the heavy lifting. I describe what I want, measure and make up ideas and learn a little terminology (fillet, chamfer) and then the machine builds me the concept that I then print and deploy into action[0].
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...
I agree with this to a large extent. This new technology is lowering the barrier to meaningfully work with existing tools. It's increasing the magnitude of jostling that may very well bump us out of a local maximum sooner than before.
I've recently been getting into musics an synthesizers and VSTs.
If you can write a program that produces a convincing piano sound people will pay you $100 per license, or more. Same thing for a violin. All you have to do is make it conform to the VST API specification and provide a bare bones UI.
But, where are they? Where are the new VSTs? I haven't seen any. Not at an accelerated rate at least.
If you can write a new and better geometry kernel for CAD software people would be willing to pay thousands of dollars. You're not building the next Google with massive distributed systems here. It's just math. To build a geometry kernal you just have to do the right calculations on a local CPU, easy. But where are the new geometry kernels? Who is vibe coding a geometry kernel?
Truck is getting a lot more work done because of AI, I believe.
But geometry kernels are severely difficult work, the maths is hard, the calculations are hard, efficiency is hard, and also policy is hard; decisions around topological naming, around the correctness of fillets (which major CAD packages do not agree on), etc.
They are among the most difficult things people have ever done with computers, which is why there are only about four of them, and they do actually put assembling large distributed platforms in the shade.
Now, will LLMs be able to help with that maths? I hope so.
Re: VSTs, I don't know much about them. The pace of development of AUv3 plugins seems to have increased, though.
I personally think the agents and the harnesses themselves are pretty cool software. Claude code, openclaw, hermes, and now the consumer facing ones. I have been doing all kinds of cool things with them. If you gave somebody the claude desktop app with blender mcp 10 years ago their minds would be absolutely blown.
A lot of these harnesses and mcps themselves are built with llms.
"Printing parts for your 3D printer" vibes. What about things that aren't for the printer?
Developing with LLMs is somewhat more expensive than without. That must be a blocker for many hobbyist.
On the other hand, corporations don't spend much on open source.
> But where is the new cool software ?
No matter what you think about it, it's impressive what the folk at Omarchy[1] have been doing with LLMs.
1: https://omarchy.org/
No it’s not. There’s nothing technically impressive about this.
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.
I use LLMs but this is probably the worst example to bring up, because there is nothing special about this as another Linux distro versus actually interesting projects like decomps and recomps.
what interesting things have they done other than repackage arch with someone else's dotfiles?
You can start with this: https://plugins.omarchy.org
Then read Omarchy news.
They made a website to download plugins?
We're pretty much still building the tools to build the software renaissance.
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".
I thought about doing that with Todoist.. and then just decided I'd rather pay the $4-5/mo or whatever to not have to use unpolished apps and sync server
I keep paying for Obsidian for the same reason. It's central to the way I work and has an ecosystem of stuff-that-understands-Obsidian making it useful.
But, for example, I did replace IRCCloud with my own custom setup. Been daily driving it for ... three months now? Took me two to three evenings to get an MVP and 3 months of iterating while using it to get it Just The Way I Like It.
I was also a paying customer of Hazel[0] for two major versions, I didn't use the majority of features so I made my own that does the things from Hazel I needed plus a few extra things that nobody else is going to care about but make my life easier.
All of these are just for me, they are open source with a permissive license, but I have zero intention of making a "community" for any of them or trying to make them into commercial offerings.
They do what I want and if someone likes them as-is, they're free to use them. If they need something else - that's why we can fork repositories =)
[0] https://www.noodlesoft.com
That's funny, I felt the exact opposite and found it fun to research how to make things sync properly and architect the solution for my LLM to then build. Currently I'm using Loro for the syncing.
Why would you do that though if 10% of the other software is better?
Maybe it's not better. Maybe your vibe coded solution is much better for your specific use case. For example I recently wanted something that could track my watch history for shows I have locally and sync them to all devices, sort of like Jellyfin, and I couldn't find something that did what I needed so I had to make my own.
Depends on the price. There are many applications that people and small/medium companies pay hundreds of euros a month for and only use a few features.
At that scale there are cases where it's worth it to create an AI assisted bespoke version. I've done two of these at $dayjob and a bunch for my own purposes. Either replacing subscription software or highly specialized versions of pay-once software.
Perhaps BigCorp is only just starting to use agentic dev properly. Plus they’ll have loads of bureaucracy still to get code out of the door. Might take another year or two.
Very simple - you don't see it because it does not get released. What you see get released, though, most people would not register so fast on their radars. And it is very bold to say "no, I don't see how antirez is puttin LLMs to work" or something along the line. Well, very often you don't see most of such people's work until it's literally everywhere. Llama.cpp a good example of such software, and yes, it accepts LLM-assisted code.
> Reverse Engineering became much easier, it is plus. Finding vulnerabilities also became much easier, another plus. Models are getting better.
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.
Do you count ArtCraft?
"that Vibed coded replacement is trash IMO" is referencing ArtCraft.
I don't get it. Like obviously any in development software will have bugs and missing features, and when it's replacing something as complex as Photoshop you can expect more issues compared to a simple text editor.
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.
> No new utilities that are super useful in day work. thats the renaissance for you - you no longer need the utilities, just ask the agent to do that, whatever it might be.
Of course you need utilities, and the AIs do too. Compare:
"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
building great software has not been constrained on throughput nor technical for a while. it's limited by the human aspect, and there is no reason imo to believe by default that ai would change this.
Yep, creative people are rare. And AI neuters the biggest driving factor of creativity, so it could backfire in the long run.
Creative people use it like a tool as with anything else.
nah, it doesn't make uncreative people any more creative but it can unlock creative people constrained by technical ability or throughput
I don't want to toot my own horn but I developed a mix of Photoshop / Pixelmator for AI assisted Photo editing and I use it every day instead of Pixelmator Pro [ 1 ]. I think it is in many respects the better app because I added many features I always wanted to have (replicate layers, any content as masks, crazy effects, etc). Granted it is buggy and still needs polish, but most software has bugs. So I've completely switched over to a complex AI generated app.
[ 1 ] https://pixellect.app/
Nobody is going to buy your AI vide coded slop.
Which part of "I use it every day" did you not understand? Its for me. I just decided to also make it available for others. The app is free.
Not sure about that, maybe not this specific software but lots of people buying AI coded products these days.
What a great phrase!
>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.
This is not representative of what's actually going on, though. It's much more like an artisan carpenter who wants to build houses using bespoke hand-tools and precise joinery. That's great, and there are beautiful works of art created by those skills. But in the last year robots were developed that build homes exactly to spec. The robots work 24/7 and finish entire houses in a few days. That same skilled carpenter is being hired to monitor the robots and check over their work. There is, economically speaking, no place in the process for him to apply his craft. In the time he rips a single board, the robots build the wall. In the time he drives a handful of nails, the drywall is up and the paint is drying. He costs $200/hr, the robots cost $1/hr.
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.
This analogy hinges on the economic argument (that a robot costs 200x less than a human), but far from being an established fact, that notion is more or less entirely based on vague, misleading and exaggerated claims from the token peddlers.
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.
> Its extremely naive, though, to simply buy into the PR from the peddlers
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.
I think the OP was talking about robotics specifically and not tasks like the one you described. Nobody can argue that for tasks like your binder replacement cost savings are enormous.
No, they were carrying on the analogy about artisanal home builders being replaced by mythical robots that do it perfectly.
In this analogy, we human software devs are the artisanal homebuilders and the fancy LLMs are the robots.
And yes, there's quite a cost to build all the datacenters for all the AI that Big Tech believes that people will want. But as you mention, the cost savings are enormous for even despite that, and will make a much larger set of problems accessible to real software solutions, that previously either stayed on paper, in peoples' heads, or maybe an Excel sheet somewhere.
It doesn't seem like they're talking about robotics, it does seem like they're talking about the AI data center build out and thus LLMs specifically.
> We can talk for ages about the quality of the robot's work, the maintainability of the resulting home
All of this stuff is so new, and we're supposed to hand wave away quality and maintainability?
Almost anyone will accept a 50% quality decrease for a 99% cost decrease. The only reasons I'm not living in a stack of cardboard boxes for $1 rent are that it's illegal, and the police won't protect it from break-ins.
> The only reasons I'm not living in a stack of cardboard boxes for $1 rent are that it's illegal
My initial reaction to reading this is not "oh great analogy", it's "please have higher standards for your quality of living"
With the money you wouldn't be paying in rent, you could build a heck of a lot of stuff. Cardboard box is a metaphorical starting point. You could build some kind of little wooden box vehicle and travel the country. If you stayed in town you could have takeout for every meal too and still have money left over. The point was not paying rent to the corrupt system, the point wasn't the exact living specifications.
You can go and live in cardboard boxes right now for free. Why aren't you? Being technically legal or not is besides the point - there are places where it is tolerated.
>You can go and live in cardboard boxes right now for free. Why aren't you?
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.
It's not tolerated anywhere I've been. I'd go to prison.
You're thinking with your engineering brain.
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.
It's a good thing structural and civil engineers don't design infrastructure this way.
Not that tired old parable again. It doesn't hold.
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.
> Robots are deterministic
iunno about that boss...
> But in the last year robots were developed that build homes exactly to spec.
That is not what's happened at all. That's the entire point of the discussion. The robots aren't as good as humans, they can't follow specs or instuctions as well, and their output is markedly worse than human output in a lot of ways.
The only difference is that the robots are faster and don't need rest. And we're discovering now that that is not good enough.
Good luck with the eventual rugpull. It always happens.
(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.
The difference is the open and self-hosted models are not far behind the closed models. You can't rugpull if you don't own the rug.
Unless they manage regulatory capture and with this kind of money at stake I would not bet against that. It's going to get very ugly. Read up on transhumanism to see what the stakes are, they're playing for keeps here.
Regulatory capture - I tried to bring this up in another comment of mine, but not much resonance.
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?
Like this:
https://madrobot.blog/2026/10/19/us-bans-deepseek-federal-ag...
And when it starts, it looks a lot like this:
<https://chinaselectcommittee.house.gov/media/letters/moolena...>
> How would regulatory capture of LLM realistically look like? Do you have any ideas here?
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.
If it's open weight it doesn't really matter if there's regulatory capture, it'll be like piracy is currently.
> Read up on transhumanism to see what the stakes are...
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 need not work, need not get sick, need not die[.]
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?
> One of these is not like the others.
Nah. All three are covered by my original statement:
Work, sickness, and death are -for the most part- all things that are awful. As it turns out, there are many more than three awful aspects to being a present-day human. For brevity's sake, I picked three that seemed to me to be most likely to be widely understood and widely considered to be awful.> 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_.".
Yes, and communism is great too [1].
[1] in theory.
And anarchy, yeah.
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.
Except the robots actually cost far more than $1/hour, it’s just that the owners have enough capital to operate at a massive loss until the carpenters have largely been replaced. Only then will we learn what the robots actually cost.
Plus robots (software and hardware) will increasingly be used to invade even the most basic privacy, to control people etc. They are not just productivity tools anymore, they are also mass surveillance tools and every bad thing that comes with it.
Example: https://m.youtube.com/watch?v=ca7ufSI1Qxg&ra=m
Most depressing part is humans are training robots to replace themselves
Self-hosted and/or open models are fast approaching (or have reached) the quality bar necessary to handle a great deal of software development work. The models needed to cause the OP's problem are becoming commoditized.
What do you run an open model on when Nvidia stops selling you compute?
Forget robots, what you're describing already happened in the trades. Modern construction techniques are all about lowering the skill required to build a house and standardizing the process so it can be done with minimal training. E.g. drywall and caulk replacing plaster and precise joinery.
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 trick to sleeping in a shit-covered bed is to slap a big sheet of plastic on it and never look under it.
The Claude Code development team has pioneered that new way of sleeping.
I have been told that it keeps you warm and adapts to your body's shape for increased comfort.
I have been told it is the future of sleeping, and if I don‘t sleep with a plastic cover directing the pooping agent I will fall behind and eventually become obsolete. And that sleeping like this will eventually cure cancer and build swarms of mining rigs in space.
Still have to deal with the stench. But nothing a nose clip can't solve, I reckon?
"shit fast and break things!"
Shit fast and soil things.
Quite profound.
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.
100%. Although large handwritten software code bases are often indecipherable to the average human being (actually, almost always indecipherable to the average human being, but often indecipherable even to skilled software engineers too); sometimes due to inherent complexities, sometimes due to Enterprise-isms.
Yeah, but only LLMs shit the bed at computer speed.
Let's be honest though, maybe "in the same way it's no longer feasible for tailors to spin their own yarn and weave their own cloth".
Tailors didn't do those things.
Those slippery slope arguments people make assume that once a technology introduces a specific risk, society lacks the agency to create counter-measures, new norms, or alternative uses for that technology.
History shows that when a slope appears, regulation steps in, technology evolves to solve the problem, or the culture shifts to reinterpret the tool.
In almost every case, the feared bottom of the slope was never reached because humans constantly built ramps or bridges along the way.
sounds like bullshit. I'm actually glad that the people that hold this view are getting out of the way, they are the people that gatekept open source and made it annoying to work with. Do I miss having to wait for months to fix a trivial bug because I first need to convince 15 people that the fix is even necessary?
No, not really.
i'm sick of scatological metaphor.
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.
Doesn't that imply that there's nothing inherently wrong about shitting in someone's bed and then sleeping in it?
What kind of autism is it that makes someone dig into the specific technical implications of a joke about sleeping in a bed someone took a shit in?
I thought Rich Hickey's sand cookie analogy was more apt than this bed shitting one.
They mean it's technically possible, but you wouldn't want to.
Society is pretty tolerant about this though?
Only if you're into vibe coding.
I was hugely into Ruby back in the pre-Rails era.
Do you mind telling us what you've pivoted to as a kind of replacement?
Who is anyone to judge another person's kinks?
I have a strange feeling that we'll see an increasing number of these to the point where we'll start to see legitimate issues in ecosystems as real talent exits (due to projects no longer evolving, being deprecated, etc as a response to AI).
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.
[1] https://www.youtube.com/watch?v=NI6xffJn4-w
Well, people aren’t coding to keep their skills and knowledge fresh.
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?
I tried to write a little project the way an AI would, going gung-ho and not stopping to ask if this is a good design. It actually went okay and now I have a silly thing that wouldn't exist otherwise. It was still slower than an AI, but only took a few days and now I know how the thing works inside, which I wouldn't if I used AI. I don't think programmers can't match AI, I think we don't want the associated responsibility of knowing we might be building a pile of shit. I think we want to avoid writing a pile of shit so we slow way the heck down.
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.
I think some people have a tendency to forget that you can half-ass things because some things only actually need half of an ass (actually a surprising number of things, though it can be difficult to predict that completely accurately!).
I've yet to see a well-loved widely-used vibe-coded (rather than just AI-assisted) app that has gone live, and then gone through multiple feature additions and remained secure and used.
"Never half-ass two things. Whole-ass one thing"
Ron Swanson
This is exactly how people were told to build MVPs for their startups, I'd have expected HN, being on YC, of all places to understand this, so it's good that you're keeping the culture alive.
> 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?
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].
[1] https://lattice.engineering
> when there's a noticeable decline in stability and development
When?
It's already started imo. Spotify's desktop app on MacOS is becoming unusable (up to ~30s delays after clicking a button before the action runs—used to be snappy). I just saw the Epic Games launcher thing. Anecdotally, countless websites just break at random (more so than even when it was 100% human).
Spotify doesn't care about how fast their app is because for a lot of their users there is no alternative. Everything else is paid only (Apple), has a bad quality library with duplicates and trash metadata (YT) or is a lot less convenient (physical media and most pirate options).
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.
The current macOS itself would suffice.
At least a year ago
It will happen, I fear, but I suspect won't be funny.
Presumably, private companies and individuals with the money to pay for tokens continue forward rolling their own in-house solutions and slowly abandon OSS as it rots. What happens when each organization has its own entire stack though? Sure, FAANGs already were like that but most smaller companies were sort of "speaking the same language" and reusing software like apache, mongodb, nginx etc.
> What happens when each organization has its own entire stack though?
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 have a different sentiment than most of what is shared here. I do enjoy coding. I do enjoy solving problems, but I feel like most of the work in my corporate/startup job life always have quite a significant amount repetitive and boring work (e.g. updating library for security reason, adding new CRUD endpoints, updating the app as per latest user study, etc). And so now I feel liberated because I can focus on problem solving instead of those tedious work. Given a problem, I can focus on thinking how to solve the problem instead of spending time coding up the solution. I don't blindly tell AI the problem and have it come up with a solution. Instead, I am particular in how I want problem to be solved, and then I will tell AI to implement it. And so in a way, now I have a 24/7 junior engineers that would get things done very quickly according to how I want the work to be done.
It's true that my ability to code would have atrophied, but I think it would happen anyway when one gets promoted and have to spend less and less time coding and more architecture or reviewing their juniors' work.
I still review the code created by AI very closely, so I don't feel like my ability to code degrades as much. Yeah, and when I feel like I have time, I would like to work on the code myself. Plus, if I want to, I can always code things by hand in my own personal project, for example.
One of the most underrated abilities in this industry is knowing when to walk away from something that no longer bring you joy. It's sad that it has come to this, but I'm sure the OP will find something more fun to do with their time.
I'd say just knowing when to walk away in anything, not just this industry, is one of the best skills you can have as a human.
In general, figuring out what you want and working towards that is much harder than people realize.
And sticking with it until that time. Both important skills
This industry is actively working on making people walk away regardless of joy status.
This hits too close to home.
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 have wondered where we are with stuff like this.
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.
Eventually maintainers will just throw up their hands and leave the repo under the care of a LLM. "Merge PRs that align with the product's vision/design/intention. Meticulously examine for malicious contributions. Email me if there are doubts."
The top tier LLMs would be as good at this as a lot of maintainers, right?
> 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.
yeah maybe. i think it's become a lot more feasible to fork and modify vs wait for new features or fix bugs. the future might be that the main repo for a project forms the "core", and consumers use their own forks which are automatically kept up to date. the core doesn't need to be as configurable or flexible as the forks can add functionality directly. in theory it could be easier for the maintainers.
I feel sad about this passing of the age of programmer because programming was a good way to make it big when you have nothing. You start your life in a poor family, you fall in love with technology, study hard and make a good life for yourself and your family by becoming a programmer. Now this career avenue (like many others) is closing because of AI. Everyone becomes less valuable and less happy.
> You want changes? Go ask your agent or something. I’m told that’s what you’re supposed to do now. I’m out.
Totally fair. The source code is open. Thanks for all the fish.
The source is open, but the ownership of the "text" and 130 other gems still resides with the blog author... maybe he could add the projects to https://www.codeshelter.co/ or something.
Can’t LLMs make forks as needed?
The LLM can maintain it now.
No need to vice signal that hard.
Seems I wrote the last non-AI book about Ruby https://leanpub.com/rubyisforfun (free).
It's definitely a great language to have fun with. I enjoyed writing the book. Moreover, I wrote it for my son who was 5 years old back then, in hopes it will be useful over time, pretty much like books about bash/sh were useful for decades.
I still remember those days I was working on my book to and from work. Those memories and dreams are priceless.
I haven't updated the book intentionally to keep the "Last updated on 2022-05-26". Was it the last Ruby book written without AI?
Anyway, the joy is over. So sad.
I can't speak to Ruby or other programming languages, but as a creator and occasional hacker with a media background, I have found AI to be incredibly liberating.
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.
There’s lots of ways to find joy in things.
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.
> 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.
That factory produced chair is way more likely to end up as landfill than a hand-made one. The AI-generated slopware generates the landfill (ie pollution and the jobkilling) before it's delivered to you.
Eeeeh, closer to 3d printing.
You can completely clone board games and steal from their original makers OR you can make something new that didn’t exist before because there’s tools to facilitate your work that you didn’t have until now
I notice you did not write the sanding phase - because no one likes that part.
Thank you for this comment.
Transcription is one legitimate use for LLMs that archivists are excited about. It's a case of a tedious task that frees them up to do the more interesting parts of the jobs, and the product is something that can be easily compared to the original and verified. This is very different from something like writing prose or code.
It definitely varies person-by-person. People who enjoyed l33tcoding, the act of programming, and the building process feel one way vs. the people who are in it for the end product, or enjoy the design process feel another. In a way, it's almost like those who enjoy the journey vs. the result (although the design process is also a journey).
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.
One thing I always like about coding is that when something is annoying or hard to do, you can improve things so that it becomes easier and more straightforward, at least in principle. For example if a library is badly designed, buggy, or just doesn’t exist, you can write your own. If tooling is bad, you can improve it or create your own tooling. At the extreme, you can create your own programming language, database or operating system. In the small, you can write your own macros and wrapper functions.
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.
Actually nobody is preventing you from building library with LLM.
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.
I think you missed my point. Yes I can work on a library with an LLM, but I can’t work on the LLM itself, which is the interface I’m corresponding with now, where previously it was the library. Previously, the thing I was interacting with was malleable in a highly well-defined way. Now it isn’t.
Yeah: in a world where I am coding, I can put in the effort to make the software work the way I think... to use an LLM, I largely have to think the way it wants to work / converse.
I get it but not every editor/IDE was vim/emacs. Some tools were more or less rigid and today you still can have malleable workflows through custom workflows and harnesses.
It's not like you need to limit yourself to Anthropic and Claude Code. I myself developed "rik" a harness that can modify only a single file and hook up blazingly fast models for it to be like wide-context autocompletion.
The malleability is still here, just different.
https://github.com/exlee/rik
1) I really wasn't talking about an editor at all: I was talking about the libraries and frameworks you use to program, so that the way you interact with the problem space is the way you think. I mean, even as someone who has almost exclusively used vim for the past 25 years of coding, if I went back to the prior 10 years before THAT of using Visual Studio, I'd have described this no differently.
2) In fact I have never used off-the-shelf coding assistants (Claude Code, etc.) and have only used my own harness; yet to me the issue is that the model itself often likes to think about problems in ways I find frustrating and unhelpful. The issue isn't at the level of how diffs are applied (that's commodity) or how context is managed (I can fix that) but in how problems are described, how abstractions are managed, and the language used to discuss them.
So many people here are talking about not coding anymore. At the same time, when a project is coded by AI, it's mocked here as being vibe-coded, and many people say they won't use it.
There seems to be an inconsistency between those facts that would be worth discussing. I'd be interested to hear anyone's thoughts.
Those two things are not mutually exclusive or inconsistent at all.
How do you resolve it? Do you think people are creating software they wouldn't use, perhaps because they are forced to create what they would judge as crappy AI software on their jobs?
Ah... maybe I misstated my question. There are people who are stopping working on FOSS projects because that's no longer fun. But what I meant to focus on is people who say they are still creating software, using AI, and therefore miss coding. And the fact that there are people who say they wouldn't use a vibe-coded project.
If those are overlapping groups, there would seem to be an inconsistency, but maybe it's resolved as I mentioned in the first paragraph of this comment.
I'm still interested in anyone's thoughts.
Some people are producing vibe coded garbage with LLMs and some people are protesting and getting out of the business. It doesn't have to be the same people.
But yeah, people who are doing both could be forced to use LLMs at work and unable to leave. And when those LLMs inevitably fail because the codebase is too complex they have to refactor with LLMs because what they originally crapped out is incomprehensible and no one bothered to read and understand it while it was being developed. I expect a lot of people don't like that we are becoming dependent on funnelling money to a third party to maintain the crappy software they enable to begin with.
I think the entire industry is about to have another recognition about the importance of quality in software development. Just like they did with lines of code as a metric, and outsourcing with minimal communication.
I'm so conflicted about this. I do love being able to create and bring ideas to life at the light of speed. But I also see the unique value in the experience of coding, deeply caring about the craft. Can we have both?
> I'm so conflicted about this. I do love being able to create and bring ideas to life at the light of speed. But I also see the unique value in the experience of coding, deeply caring about the craft. Can we have both?
Yes, but only if you're very wealthy.
As always, everyone else can suck it.
I'm working two entry level jobs, and I have spent maybe upwards of 100B tokens in the last year? on a project which isn't more than 250,000 lines of code. I have done countless refactors and care a lot about the craft.
Managing a codebase that big is not an entry level job. Can you outline very roughly what those projects do please?
I work at a dentist office and a steakhouse I built my lisp machine interpreter web framework on my own time
These sound great! Do you have any links for that web framework? What was the biggest challenge in that project so far?
https://pricklypear.rocks/
Performance is a consistent issue. Ive had to build a lot of optimization and benchmarking tools in order to keep things functional. Every couple of weeks I have to refactor everything to keep load times good. Both on the database side and on the lisp side. The models, by default, will happily build functions with 26 database calls or millions of recursive steps.
More frustrating, though, was probably the 3-4 weeks when i migrated the storage concept to "one big tree". This was an aesthetic thing that I later found out some of the old lisp machines also did. But - I didn't think about it initially when I built it. And switching things at that low level was difficult. There was a lot of legacy data stores along the way and programs were using the old store constantly. Very irritating. I still wonder if it would have been faster to just completely rebuild at that point.
An ambitious user could probably run it themselves and have it function, or I think I could run it as a managed service for somebody and have it be good, but the UX is still not really ready for primetime
I think it's like any other craft. Buildings and most furniture are built using mass produced lumber and machinery, but many people still enjoy doing woodworking by hand.
Yes, but it may be hard to justify the "experience of coding" from a commercial standpoint.
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.
No.
If the client cares for your craftsmanship.
You can have both, but there are very few people who care about it purely as a craft. They want pay, recognition, etc along with it.
As a hobby you can’t live from
Yes, choose both or choose one of them and, for god's sake, STFU and make something!
"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."
yup. agree.
Well, it might be a rational view for the author, but for me that's like saying "I no longer like cooking, because there are companies providing the recipes I like as ready-made meals", or "I no longer play my instruments and make music, because there are machines which can create the music I like"; both topic apply to me personally, and I wouldn't agree to the statements. I still like to design and implement systems, make clever architectures and solve tricky problems. Why should I care if other people use LLMs to do so?
For most of us, as much as we enjoy making clever things, without the context of a job there aren't that many fun problems we have time to engage with.
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.
I'm still in business myself and I see the problems. But the article was not about what you suggest. It was about contributing open-source to a community and whether it is still worthwhile and fun, even if some community exponents started to use LLMs.
>Why should I care if other people use LLMs to do so?
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.
Ruby is still fun to write, in fact I make the LLMs write the underlying libraries in Rust while I enjoy writing Ruby on top. win win
Parasites that kill the host are bad parasites.
I felt inspired by this, and removed all my Github projects (except a few that people may depend on, which I just archived).
If AI can do what we deem creative work, say implementing a chunk of a Ruby gem, should we still consider such work creative in the first place? I was wondering if we should face an uncomfortable realization: much of what we call being creative was really the cost of overcoming cognitive friction due to the limit of our brains.
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.
> If AI can do what we deem creative work, say implementing a chunk of a Ruby gem, should we still consider such work creative in the first place?
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.
Yes. being good enough alone can destroy lots of job opportunities, unfortunately.
I think this is really an excellent comment because it challenges us to reframe what we take for granted as creativity, enforces us to reconsider what creativity could be in the face of really powerful automation.
I feel the same, AI driven programming has killed the joy, now I rather watch some TV show.
That's a real issue in my opinion.
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.
Ruby seems like one of the worst language for LLM development for all the reasons that some humans tend to like it.
Is it? I thought I read somewhere that it was one of the most token-efficient ones. Also, Rails + LLMs is magic and speed (of development).
Ruby is token efficient because it is terse, and has little special-characters. That why it sometimes "reads like english".
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.
This is going to be happening all over the place. I've released some code in the past but I'm definitely not going to do so again without ironclad agreements in place about AI ingestion. I'm not going to help the likes of Amodei and Altman to a further increase in their power, my goal was to help normal people not weird cult-like corporations and I suspect that a lot of people that have released code over the years have the same attitude (though obviously not all of them).
Yes when it becomes clear that half the participants aren't playing by the same rules you were its only fair to flip the table and walk away.
I'm no longer motivated to share when that only ends up feeding the machine intent on replacing me and I'm no longer interested in looking at other's contributions when it likely contains vomit from that very same machine, whether disclosed or not. This doesn't bode well for the future of open source or free culture in general and I do hope that well get the societal and legislative changes needed to make sharing a remotely fair proposition again but, considering that the entire economy has been essentially a big casino for over a decade now and all the winners are not held accountable for the issues they cause because doing so would bring the entire system crashing down, I'm not optimistic about that happening within my lifetime.
I think there is some room here for an update to the GPL. GPL V4 that is AI proof would be a big milestone.
> LLMs need to consume more good human-written Ruby code. You might already have seen the problem here.
You might already have _been_ the problem here.
Fixed that typo ;)
His post made him seem like a nutter and then I read his main blog and ya, he's a nutter.
> You want changes? Go ask your agent or something. I’m told that’s what you’re supposed to do now.
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 feel like this is valuable context to why a Ruby dev may feel disheartened https://youtu.be/IEOqgPfGLQ8
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.
Why wouldn’t the rails/ruby users just use rust too?
In short, because DHH hates programming in Rust.
> In that context, I no longer feel inspired to contribute to a commons that’s being strip-mined by massive corporations
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
I am pretty sure that the whole open source ideology was developed and disseminated exactly by a cartel of those massive corporations with the sole intention to amass as much human-written code as possible to train AI.
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 really would like to believe that GPL is the answer here but I know it won’t change anything for them. If they can steal and destroy books, nothing will ever stop them.
We need to go beyond GPL and have the government invest in public software programs.
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.
If LLMs aren't an example of trickle down, I don't know what is
Paying for something that has been scraped, trained on hundreds of billions of lines of code (mixed up GPL and other licenses) and being resold back to you is trickle down?
Paying 20$ per month to double my productivity, answer many of my health questions with good accuracy, help me with DIY projects around the house and do almost as good of a job as a psychotherapist is trickle down for me.
And you don't even have to pay - there are perfectly fine free tiers and open weight models.
> whilst making billions for their shareholders
> I am yet to see this trickle down
hmm...these two sentences don't agree with each other.
FSF has been the greatest pro-corporate activist, convincing thousands if not millions of programmers to give away their code.
Is it me or do most of these "I'm quitting programming" posts seem to be coming from the Ruby community?
It would not be absurd: Ruby (not Rails, Ruby, I insist) puts very high the "joy of coding", its syntax pleasant to write and read for humans, and its welcoming community. It never was the most performant, it's been a long time since it last was the cool kid on the block, but all those other aspects make it very attractive for very different reasons. That's a big part of why I started writing Ruby a long time ago vs. other languages, and why I enjoyed it so much, and I don't think I'm the only one.
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.
i saw those dollar signs and @ symbols in 2003 and ran screaming, myself (coming from perl, seeking to...get away from it)
DHH hasn't been helping.
yes, a lot of them hate dhh and believe him to be problematic, and since he's a big ai guy now the natural thing is to oppose him on that
Ruby is definitely not bringing me joy since I have to see needlessly depressing articles like this. Rails did not bring me joy 12 years ago when I needed to update dependencies for a legacy cloud service we relied on. Ruby's middling syntax didn't bring me joy either, despite the hype of the time. My child named our cat Ruby, and I had to explain to fellow nerds that it wasn't inspired by the language.
I find it amusing to hear these anti-dhh rants. These old schoolers should remember the old “Rails is a Ghetto” rant. It’s got much of the same technical criticism in there. DHH was never a star, but he’s good at marketing.
I really enjoy Ruby, rails, and agent-driven development. So I’m probably not the target audience. To each their own.
The robots aren't as good as you think. Its just that we're sceamed at all fay by hoardes of people saying they are.
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.
But the robots are mostly "fine", and "fine" is good enough for most things we get paid for: enterprise SaaS products, mobile apps, etc.
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.
This. Enterprise software is going to get a lot worse.
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.
> but the money's still decent
Do you realize it won't last long at such rate?
That's a fear I have, yes. Not much we can do about it, though. We've collectively been pretty lucky so far, in terms of compensation. Can't last forever, one way or another.
Delusional take. Vibecoders love to roll around the internet and tell SWEs that they're about to lose their livlihood, its a weird behavior.
I'm not a vibecoder, I'm an anti-one, won't touch that shit with a 10' pole.
I know of at least 3 significant rails projects/donors that are actively discussing AI-lead refactors since DHH's talk declaring rails dead. Even if you like the language it's going to be a total mess in a year or two.
I'd rather tell LLM to rewrite project in Python. There is no good reason to use Ruby/Rails now, the community and job market is shrinking and the language is nothing special.
I would take Ruby over Python any day of the week and I find Ruby very flexible and elegant. It’s all personal preference.
I’m in it for the love of creating software and making computers do things not possible before.
It’s never been a more exciting time for software development.
Perhaps never been a less exciting time for handcrafting code.
"But even beyond Rails, the Ruby language itself is now substantially developed using LLM agents. Matz is giving talks repositioning Ruby as a “Programming Language for AI age”, to be emitted by LLM agents, and stating that to be able to do that LLMs need to consume more good human-written Ruby code. You might already have seen the problem here."
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.
> The biggest LLM sellouts in open source
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.
> Be and let be
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".
Why would it be wrong to call Linus a sellout? The guy started with a hobby project that eventually became big, was mostly taken over by corporations, and he is now doing mostly corporate friendly things.
Yep Linus certainly jumped the shark at the 2026 maintainers summit the other week.
> Ironically Oracle [...] is now being the good guy.
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.
You can't both be nice and believe that LLMs are useful tools?
Did you read the quote from the article?
Seems unnecessarily harsh. LLMs are a power tool. I'm not on-board with how it's being used in a lot of places, once things shake out I believe it's going to be net positive.
'Oracle is the good guy' is a sort of canary in the coal mine that perhaps you've got your worldview a bit upside down
I'm glad someone else called out Matz on this.
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.
Yep, the uncultivated race car driver (dhh) and the snowboarder (Lütke) own Matz now. They can insult him in public and he complies.
I'm not sure why he isn't rich enough to walk away.
So many questions!
What exactly is an LLM sellout? Someone who sees the world moving toward it?
Wouldn't Rust be a sellout too?
Why is Scala an LLM sellout?
a LOT of these arguments against New Thing seem to always fall into the category of "I no longer receive social credit for being a productive member of a niche in-group, so I quit.", masked as "They peed in the pool!".
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.
The social aspect of a craft is more than just "clout" or "smart creds" from an in-group, it is also collaboration, commentary, discussion, and the joy of helping others find their feet, for example. LLM-centric programming takes away not just the "clout" part but also all the other less egocentric social aspects of programming as a hobby. If the hobby's ecosystem gets whittled down to posting a finished demo and saying "look at what Claude did," well, what is there to talk about, really? The whole "process" aspect of the hobby gets obliterated, or at least effectively ceases to exist as a social phenomenon.
This is an overreaction. Yes, with AI we can all write our own features and bug fixes for open source, but there’s still community benefit in having a canonical version somewhere, rather than everyone writing bespoke solutions for themselves.
> there’s still community benefit in having a canonical version somewhere, rather than everyone writing bespoke solutions for themselves.
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
Would you prefer we leave it entirely to the AIs? That’s an option, probably the default option. If you’re opposed to it, then why advocate for it?
I'm not really advocating for any position here - just pointing out that open-source maintainer was a thankless role before LLMs, and it's only become worse since.