5 ms·OpenAI, Anthropic, Google have been in talks on AI safety for weeks2 points by Alephinitesimal 17d ago