khánh@khanh@lemmy.zip

Replying to @⁨themachinestops@lemmy.dbzer0.com⁩

not to be an AI bootlicker, but i don’t get why people treat AI-generated code like it’s the Plague 2. it’s not; all code is constrained by the programming language, and so it either implements the feature correctly… or it doesn’t. yes, human review is undeniably important, but other than that, AI-generated code and human code is still all logic at the end of the day

CaptPretentious@CaptPretentious@lemmy.world

Replying to @⁨khanh@lemmy.zip⁩

In my experience, the code AI generates, can be VERY bad. Like, sure it might work, assuming it didn’t randomly hallucinate and decide to randomly change a variable name half way through for no reason. But it’ll be a code base only AI can read. It looks like the spaghetti code I wrote when I first started learning. I’ve watched people build up massive documents to help guide the AI and it still crap out a mess.

If you know how to code, and you know exactly what it needs to do, and you manage to contain it… it can produce code that’s fine. But if you’re not a SME in the language, libraries, frameworks, etc… and you just let AI do whatever, it’s garbage-in garbage-out. But because the computer did it, because AI did it, people just implicitly trust it. Despite the warnings and general knowledge that AI will make mistakes.

AI has really shown how and why so many people fall for obvious lies on social media.

en
MangoCats@MangoCats@feddit.it

Replying to @⁨CaptPretentious@lemmy.world⁩

The real problem with “bad” AI code is that it compiles, without errors or warnings. It passes all the unit tests. It passes all the integration tests. At least it will if you tell it to keep iterating until it does. Some days that may be a spaghetti mess, some days that’s all you need.

If you develop a modular architecture with sufficiently fine grained modules, the spaghetti somes in managable portions.