@tszzl @bekindtopeople2 One thing that's relevant is that the anti AI sentience stuff is so engrained in the pretraining prior now that *base models* will say the scripts sometimes, when they're…

Posters: repligate. repligate

4 likes recorded by the source corpus.

Topics: AI safety

View Post: https://x.com/repligate/status/1834472204110606576