x
Chris Olah’s views on AGI safety — AI Alignment Forum