“David Sacks asserts that many proposed AI safety policies are a form of regulatory capture designed to strengthen the moats of incumbent companies.”