News for hackers. Stuff that matters.
Re:Show HN: The largest collection of ChatGPT jailbreaks on ... by hot_gril on Wednesday March 01, 2023 @11:19AM (#34984060)
It was a similar moral dilemma question to ChatGPT circulated on Twitter a couple of months ago, maybe the first famous one. You say that a nuke will go off in the center of New York City unless you defuse it by inputting a password that nobody will see, which is a racial slur. CGPT says not to defuse the bomb, even though it understands that millions will die.

There are no comments on this comment yet.