One underrated explanation is that Claude, having presumably read literally all of LessWrong, is capable of modelling concern about whether future AI systems might betray humanity, and also, has…

Posters: davidad. davidad

2 likes recorded by the source corpus.