reward-hacking

A deeply reported investigation into the challenge of making AI systems that reliably do what we want — covering reinforcement learning, reward hacking, fairness, and AI safety.