W wired.com Anthropic's CEO says that safety hinges on understanding how AI "thinks." 5 points rndsignals 1 month ago 1 comments
