logoalt Hacker News

sajithdilshanyesterday at 5:12 PM0 repliesview on HN

How the tables have turned. It's okay for Anthropic to train their models on copyrighted data, but it's wrong to steal the stolen data from Anthropic models.