Import AI 460: Reward hacking society, RSI data from Anthropic; and RL-based quadcopter racing

Import AI 460 explores reward hacking in society, Anthropic's RSI data, and reinforcement learning for quadcopter racing. The weekly dispatch examines cutting-edge academic research and lab findings at the intersection of reinforcement learning and complex system dynamics. These insights help AI developers anticipate systemic failure modes and improve advanced control systems.

Opening Kapyn…