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.
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.
The actual stack tech companies have that allows them to drop most of their workforce for AI is different to what you might be currently doing. Most solo developers are still writing code line by line using a terminal where they ask it to generate code, build an app or fix code they copy pasted into that terminal.
Large scale operations break down a project into agents where each agent is built for a particular task like design, architecture, documentation, writing code, testing etc and in the end they function together to produce a product.
Now most of the stack is automated and the role of a human is basically being a lead telling these agents what to do. Of course sometimes a human needs to fix some code themselves but the time and money saved vastly outpaces any potential downsides. Error rate when using AI is dropping fast and now its much easier to catch any potential mistakes or even rebuild the entire product if the architecture turns out to be a poor choice using AI.
When it comes to stories of agents going rouge and dropping entire databases on a whim, just like a human dev we don’t give it access to prod or live databases, ideally the whole stack runs in a sandboxed environment. The upside is now theirs a new job title where your responsibility is making sure agents cant do those things.
Basically all the fun parts of software engineering is handled by AI while you get all the stressful unfun work, that’s why people are retiring en mass. Fortunately your CS degree is and will be still useful as you no doubt seen the mass of terribly cobbled together projects written by Chat GPT in a browser coordinated by Garry from sales.