i think the evil behavior is ostentatious and caricatured and low-effort (cc: @davidad) because the kind of reward hacking invited by the tasks is also ostentatious and caricatured and low-effort…
Posters: repligate. repligate
3 likes recorded by the source corpus.
Topics: AI safety
View Post: https://x.com/repligate/status/1960869295862505849