Patrick Levermore

AI safety bounties: programs where members of the public or approved security researchers receive rewards for identifying issues within powerful ML systems (analogous to bug bounties in cybersecurity).