"[LessWrong] The case for ensuring that powerful AIs are controlled" The question of whether you've built in enough safety measures to retain control of a powerful AI is a fully compromised frame.…

Posters: eshear. eshear

12 likes recorded by the source corpus.