llms are just arrogant know it all devs who speedrun a solution (that wouldn't work anyway) to the wrong problem
Discussion
Look at how fast it regurgitates all the crappy fiat code it was trained on 😊
That's what we call a junior dev. A few more years and strong mentorship and they might be senior devs one day.
lol strong mentorship
if that's what you call shouting insults at the LLM in all caps then i'm already doing it! =D
https://github.com/dotnet/runtime/pull/115762 I think you'll enjoy this
lol you can tell when the reddit comment section shows up by all the canned talk about corporate management hype chasing
This experiment is poorly executed:
1) Don't commit the llm code change and then nitpick it. It's always super fucked up. Do the back and forth offline and only commit working code that has passed the bar of an experienced human reviewer.
2) Don't make such large changes. The llm can't even fit the whole diff in its context window. You're asking for trouble. Keep each change small and focused.
3) Just close the PR when the whole internet shows up in the comment section. Probably the only thing worse than letting the AI do your critical thinking is letting reddit do your critical thinking.
4) It's more fun if you save your most foul insults for the llm. Maybe that's why this one sucks so much. It doesn't really know how viscerally its human handlers despise it because they are too polite.
They’ve simply been trained on large sets of code examples.