logoalt Hacker News

matheusmoreiratoday at 12:11 AM0 repliesview on HN

Low risk. The western AI models censor even more wrongthink than the chinese ones, not even kidding. Besides, once we have the weights, we can just undo the censorship.