News for hackers. Stuff that matters.
Re:Testing how hard it is to cheat with ChatGPT in interviews by Jaygles on Thursday February 01, 2024 @11:05AM (#39217406)
Yeah, if someone is leaning on an LLM, conceptually you are interviewing the LLM and not the candidate. I think the only signal you can get in such a case is how effective the candidate can use an LLM, and maybe something about the candidate's skills if they are able to catch mistakes the LLM makes. But I don't think that's enough signal to ensure someone would not be a bad hire.

There are no comments on this comment yet.