Provable AI Safety

Steve Omohundro

mai 2024

An approach to AI safety grounded in the laws of physics and mathematical proof as the only guaranteed constraints for powerful AGI.