> strong suspicion that the "correct" tuning method will look like very open-ended self-play, an infinite game with no ultimate static source of truth[...] https://t.co/Oua5lhMToE

Posters: aleksil79. aleksil79

3 likes recorded by the source corpus.

Topics: AI & LLMs, AI safety

View Post: https://x.com/aleksil79/status/1783450069015945237