Reinforcement Learning's Real-World Power