How would an artificial intelligence (AI) decide what to do? One common approach in AI research is called “reinforcement learning”. Reinforcement learning gives the software a “reward” defined in some ...