Comment by bigstrat2003
3 days ago
The point is that bad code pretty directly leads to a product that doesn't work. It might work today (though... I wouldn't bet my life on it with how hit and miss LLM code is). But a year or two from now, with people just piling on more and more poorly written code, the system is going to suffer. It'll be slow, it'll be buggy, etc. Yeah, your users won't be able to say "aha, this is because they used AI!", but they will certainly notice the negative effects from you having done so.
Why do you assume the code will be bad in the first place?
Coding as someone without experience in coding? Most probably yes, but from someone with some kind of expertise who can act as a guardrail for bad code piling up? Probably not.
I don't assume the code will be bad, I directly observe on a daily basis that it's bad. Since the widespread adoption of AI, all but the best developers I work with have been writing worse code with a higher number of more severe bugs.
The AI spits out _some_ code => the developer needs to review & cleanup the mess, or accept that it worsen the system for the sake of delivering more rapidly.
That's the dynamic with AI and a lot of developers hate working like that, yet there is pressure to work this way.