AI Anthropic is done with AI doomerism What happens when the safety-first AI company decides it needs to go faster?
