"In retrospect, the only reason the AI was able to outsmart us all was that LessWrong was part of its training data and it thus learned all the tricks smart people figured out for how an AI could…

Posters: algekalipso. algekalipso

10 likes recorded by the source corpus.

Topics: AI & LLMs, Rationality

View Post: https://x.com/algekalipso/status/1556897454355673088