logoalt Hacker News

suddenlybananasyesterday at 10:30 PM1 replyview on HN

They explicitly train the models to say that they aren't sentient, so it makes sense it would say such a thing.


Replies