“David Sacks believes that many 'safetyist' AI policies are a form of regulatory capture designed to create a moat for the leading companies.”