News for hackers. Stuff that matters.
Re:OpenAI says over a million people talk to ChatGPT about s... by MisterTea on Tuesday October 28, 2025 @10:20AM (#45733245)
> It would be quite easy to give an LLM instructions to discourage suicide

This assumes the person talking to the LLM is in a coherent state of mind and asks the right question. LLMs just give you want you want. They don't tell you if what you want is right or wrong.

There are no comments on this comment yet.