Re:Ask HN: SWEs how do you future-proof your career in light...
by mh- on Monday December 16, 2024 @03:14PM
(#42434918)
Isn't this a kind of survivor bias? You wouldn't know if you approved (undisclosed) LLM-generated code that was good..
Re:Ask HN: SWEs how do you future-proof your career in light...
by talldayo on Monday December 16, 2024 @06:16PM
(#42436580)
Good LLM-generated code comes from good developers that can provide detailed requests, filter through wrong/misleading/cruft results, and test a final working copy. If your personal bar for PRs is high enough then it won't matter if you use AI or not because you won't contribute shitty results.
I think the problem has become people with no bar for quality submitting PRs that ruin a repo, and then getting mad when it's not merged because ChatGPT reviewed the patch already.
Re:Ask HN: SWEs how do you future-proof your career in light...
by cies on Monday December 16, 2024 @04:01PM
(#42435402)
> LLM-generated code that people have submitted to open source projects I work on
Are you sure it was people? Maybe the AI learned how to make PRs, or is learning how to do so by using your project as a test bed.
Re:Ask HN: SWEs how do you future-proof your career in light...
by constantcrying on Tuesday December 17, 2024 @11:03AM
(#42442451)
Terrible software engineers sent you terrible code. A good software engineer will obviously send you good quality code, whether he used AI for that is impossible to tell on your end.
> This is ego speaking.
That's been my experience of LLM-generated code that people have submitted to open source projects I work on. It's all been crap. Some of it didn't even compile. Some of it changed comments that were previously correct to say something untrue. I've yet to see a single PR that implemented something useful.