Just to make this clear up front: I was always very skeptical of LLMs (I’ll never call them AI) and still see them as fundamentally limited and unreliable. See my post history if you don’t believe me.
Having said that: I work in business software and the models have progressed to the point where typing code and writing classes and functions by hand has become uneconomical. We review the output closely to avoid the idiocy of “vibe-coding”, but the quality has improved to the point where we rarely have to outright reject solutions. The discussions have moved to high-level architectural points, where we just request better implementations than the first-shot. We get them, review again and move on. Maybe there are deeper, latent problems beeing embedded in the code-base through this process, but we can’t spot any although we are looking intently.
And just like that, my profession has died. Nobody in my team writes code anymore. Things that would have taken days are now done in minutes. We are all just considering design and architecture. We are changing our processes since they are not fit for the new way of doing things. Scrum? What’s even the point anymore? We can work through a sprint-backlog that would’ve taken two weeks in two days now. We do reviews 90% of the time. When we request changes they come back in minutes.
I have a feeling like the ground has been taken from below me. My son wanted to become a developer, that isn’t a real job anymore. Everyone around me feels the same way, including our management. We are a small team and a small company, everyone is effective and our jobs are not in danger as long as our product is succesful. But my profession is essentially dead. It’s a craft I honed for many, many years and the craft itself is now automated. I wanted to improve my typing skills to code faster, this is now a ridiculous endeavour. We don’t know what comes next, but I now only have my analytical skills and knowledge of good architecture to fall back on.
And it all changed so fast. I don’t think this will transfer to other professions so easily, but for programming LLMs ended it all in the span of the last four months. I thought I knew what my future holds as a senior dev, now I can barely see past christmas.This is a nightmare.
“Computer” used to be a profession.
Might be the same with “Programmer”. But also, it’s usually called Software engineer or architect for a reason. It’s not only programming. Like you said yourself you are still doing stuff or not? So its more like your profession fundamentally changed or to put it positively it rapidly adapted to new technological developments.
Now we all know how the typewriter repair people feel.
LLMs have made writing code more efficient which used to be the bottleneck but now reviews, requirements engineering, system analysis, architecture and all the other parts become the bottleneck.
This ties also into Jevon’s paradox which says that making a resource such as coal more efficient increases consumption. I think this very much applies to software also, there are tons of projects such as creating custom solutions for a client that have become a lot cheaper and will lead to a lot more software being written. If Jevon’s paradox holds then we will see increased demand for developers in the future.
The job is changing but there’s a very strong possibility of demand for developers to increase. Imagine someone vibe coding their way to a usable product, that guy will need to hire someone. Businesses that used to be in shopify or WordPress get an internal system for workflows bolted on or get a custom website developed.
I myself learned Vim, edited code super fast, wrote tools and abstractions that actually helped and could implement features pretty fast and that skill I honed for 10 years is just not valuable anymore. The other skills such as reading code fast and debugging has become way more valuable though.
It could very much be that now is a great time to learn programming since people are not signing up and even though right now supply is higher than demand the demand could increase substantially as more software gets written. One thing is for sure though, our profession is going to write more code and that code will need maintenance and modification.
As a (now ex) developer I feel exactly the same. And worse, I can’t even look intently at the codebase. Our management pressures us to ship as fast as possible and the original codebase was already a mess. I just prompt an AI and ship it in production.
I’m lucky I started working in the industry before AI and that I started developing in 2013
I’m gonna try to not make this a pro-AI comment, but I’m still going to try and offer you a glimpse of hope as a fellow person who used to handwrite code for a living and sometimes still does. Unlike a lot of other people using AI, I only really started using coding agents a few months ago when it was clear they’re getting pretty good.
We are all just considering design and architecture.
So you now get to think more about the what and the why, not the how. Not ideal if what you want to do is solve clever technical problems*, but on the other hand, code was always a tool to build solutions, not an end goal unto itself. In a software project, code is the beams, the nails, the planks and so on. You, an architect, design houses. The house won’t exist without the minute details, obviously, but what the customer needs is a well laid out solution that is also executed well, not just extremely well sanded supported beams.
We are changing our processes since they are not fit for the new way of doing things. Scrum? What’s even the point anymore? We can work through a sprint-backlog that would’ve taken two weeks in two days now.
Is that a bad thing? I always hated scrum. I thought everyone except team leads did.
We do reviews 90% of the time. When we request changes they come back in minutes.
That part sucks. But I feel like the more senior you get, the less you were always able to avoid reviews. As an absolute junior, I did 0 reviews. Later on, I spent an hour or 2 a day doing reviews after about 3-4 years… But I’ve also seen people with 15 YoE doing mostly code reviews all day before AI was a thing.
* There are still projects where this manual management is quite necessary. Look at ffmpeg, they use inline assembly to eek out extra performance. Super high performance projects in particular, will still need manually written code for the foreseeable future, even if AI can help you profile things quicker, etc.
I still am not a fan of relying completely on another company as we all know that companies have a habit of screwing their customers. Especially when they hold all the power. And we are currently still in the “get your customers addicted” phase.
When that ends prepare yourself for higher prices and lower quality. And they will increase prices and lower quality again and again. And so will the competition.
If you have to rely on llms you should set up something for yourself that they can’t take away, because the enshitification will come and it will hit the companies who rely on llms HARD
I’ll never call them AI
Good, because they’re SI now /s
Rest of the world: “Hey, can you finally stop using imperial and start using SI?”
Monkey’s paw curls
I work very low level embedded stuff, and over here LLM still can’t do shit, outside of some very specific applications, and everyone pro enough could do it just as fast anyway. However, I do witnessing the death of profession here too, people practically over the course of a year stopped being able to think for themselves. Most of the people are doing debugging through LLM, asking it the most benine questions.
At this point, I just wish I will die before the total collapse.I work as an embedded developer as well. I would have agreed with you some months ago. Now it can recognise a full proprietary compiler by one of its flag, which is used maybe a handful of companies. Scarry. I am 99% sure it “learned” by previous attempts by developers. Aka the AI company stole the conversations and trained the new model on them. The uC we are using is non public, only we have it, yet somehow the AI knows about it. The only thing it cannot do is grabbing an oscilloscope to debug the HW…
I feel this one. The company I work for just layed off over half the software team, from engineers to product managers, it has been brutal. I have been transitioning into architecture for a while now so I could see this coming, but it is devastating to our trade.
I still try and do some coding by hand. Agents have improved massively and they’re now really capable but I can’t see how anyone is going to fully automate the software development process. If there’s even the slightest chance that an agent can go off track and cause an outage, you’ll need a person to sort it out. If nobody can understand the black boxes that these things make when unattended, who’s fixing that shit? I’m yet to work for a company that is okay with a production build being down for any period of time.
There are companies running the narrative that their developers don’t write code anymore but unless they plan on fully automating the process and taking humans completely out of the loop, I don’t see how that is sustainable. There will be a point when developers realise they have no future at your company so they’ll leave, taking their knowledge with them. The bridge between your software and your business is now an LLM. That’s a huge amount of trust to put in something built on probability and has a limited context window.
AI isn’t disappearing but at some point there is going to be some high profile events that even the CEOs can’t ignore. Their job is money and if something in their business is losing them money, they will change it. That could be outages, customer dissatisfaction or their development ‘team’ is now costing them a fortune. It just sucks that we have ride this shit out.
I don’t know man. I’ve got a project I built by hand for years now. It’s big and complex. Around 200k lines of code. I don’t think AI could do that.
Ok, it could probably build something similar. Something that has the same features. But they wouldn’t work well.
My project is quite efficient at this point, and I’ve never experienced any AI apps even close to that. The bigger they get, the slower and buggier they tend to be.
I am curious to see if any vibe-coded app is still maintainable in 6 months.
Well-architected ones will be. One-shot apps probably not so much. You get out of it what you put into it IMO.
If you describe a good architecture to build, you’ll get a more maintainable solution. If you just say “build x don’t make mistakes”, you’ll get… something.
The other big problem is persistence of course. If you just oneshot an app and shit it out on github and never come back to it to fix bugs or whatever, then the issue isn’t the LLM output’s quality, it’s the lack of dedication. And LLMs make this so much easier, because it used to take months or years to build something useful so you HAD to be dedicated.
Saw a vibe coded game recently where the repo is just an entirely rebuilt codebase every commit/„version“. It’s both hilarious and grotesque.
Game developer here. Lots of people want to become the fabled “10x” developer (where did it even come from), but at best they’re just 1.1x developers, and at worst they’re even worse than before.
There were many deterministic tools for what AI can do. Save for some major differences: the AI doesn’t want you to spend weeks to months to learn a CLI tool, that has unintuitive key bindings, has unintuitive names for key operations; all while it can be controlled by natural language. The natural language is a trap, since it results in a very nondeterministic tool, while often the main problem is a previously elitist userbase that wants you to either never raise any question why it could get a GUI interface, or that they really gatekeep you from certain aspects to try to direct you into the right ways of using something. I got told by GDB users that I have to learn scripting, because that’s the big selling point of CLI tools, all while I couldn’t find anything about checking values in case of a break — it was
print, but they thought it was less important than writing scripts that automated stuff, or wanted me to learn an unintuitive name for a common process and then ask that. Some instead talked me about how much they loved reading manuals, and I should do the same (still didn’t find whatprintdid). This all while I could use RemedyBG on Windows, a fully GUI debugger, with everything having written onto the GUI, no lengthy manuals need to be read through to learn they call checking values as “abracadabra”.Art kind of has a similar issue. You don’t need to learn everything, however some elitists are really want you to leave certain styles behind the moment you hit your twenties. Or took the “good artists can do life paintings really well” way too seriously. And if you’re really unlucky, your main interaction with artists could be limited to these people.
They’ll be back, and begging: https://en.wikipedia.org/wiki/Ironies_of_Automation
Anyone can use an llm but the quality of the output is significantly different in the hands of someone that knows what they’re doing vs someone that doesn’t. Software development as a profession won’t die. The field and necessary training will evolve.
I could give everyone in this room a chainsaw and and hour and I bet at the end the results are wildly different.
I don’t feel this existential dread yet. I do write less code yes, but many of the best libraries, products, etc were written by hand.
Engineering is far from dead because you still need an understanding of how things come together to use AI effectively for the things you use it for. It’s just another tool.
On the brightside, for those of us that have been too tired after work to write code in our spare time now have the time to do it. Write that game. Make that lib. Nothing is stopping that.
We went from wire wraps to switches to punch cards to assembly and terminals and compilers and this abstraction or that. It’s just the next abstraction layer.
Careerwise? The biggest issue is dumb people with too much authority that think they can replace all human talent. A real risk, but I bet that will never work.
Weather the storm.
Dumb people with too much authority
So, the MBAs are now double MBAs?
As a (not so anymore) newly graduated computer engineer, this made me cry. Whenever I see threads like this the consolation thrown around is that someone with experience will always need to check the code and seniors didn’t used to write much code anyways. And juniors are all LLM jockeys anyways so who cares.
Well fuck me I guess, the last train to break into the industry departed 3 years ago when I thought I was learning my profession. A couple months ago I felt principled, I wanted a job where I wouldn’t be forced to vibe code. Now I’ll take anything.
At this rate even if I do manage to get hired, and even if by some muracle I’m allowed to write code there; the experience I gain from it wont be valued the same as someone who started working before LLMs became widespread. I’m a part of the slop generation now.
Makes me wonder what answer pr people want when they ask if and how I use ai. Do they want me to use them or not? If I use them, I accept that I have no place among the senior developers of tomorrow, if I don’t I’m missing on the one skill that companies looking to hire me expect me to have for the sake of something likely to be obsolete.
Maybe I shouldn’t have made being a developer a part of my identity while still in university. Is it time to let it go? Accept that I’ve spent years on something I wanted and believed would lead to a good life for it to just not pay off? That the safest most sensible fucking life choice imaginable just left me unemployable?
Maybe it’s for the best, as they say “Don’t turn your hobbies into a job”. But it breaks my heart to let go. Where would I even pivot here?
Sorry if this was rambling. This is a cry for help
Do they want me to use them or not? If I use them, I accept that I have no place among the senior developers of tomorrow
Untrue IMO. A senior developer knows when to use which tool. As a future senior developer you’ll need to know when to use AI and when not to.
I see people who have been writing code and architecting systems for 20 years use AI to speed up parts of their workflow. More time left over for architectural work when 2 hours of debugging a stupid bug can be a prompt fired away while you go think about more big picture things.
If I had to interview for a job right now and was asked about AI usage, I would say that I’m definitely going to lean towards more manual work in the beginning to understand the projects and the tech stack better, and once I’m more comfortable with my knowledge, I’ll use more AI again, as I can verify its output. In the meantime, it’s still a fantastic search tool too.
I’m not an expert, but maybe you can say you use LLMs for some projects and develop others by hand for the best of both worlds?



