“Leopold Aschenbrenner was fired from OpenAI for allegedly leaking a brainstorming document on AGI preparedness to three external researchers for feedback.”