Hacker News new | past | comments | ask | show | jobs | submit login

I'm sure every LLM provider has in their TOS that it should not be used to write safety critical software.

Like how itunes should not be used to design nuclear warheads.




But according to author, it is not possible to write safety critical software at all. You just can't trust code. It runs randomly after being copied from StackOverflow or ChatGPT.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: