This is a video from AI Warehouse of the process of training five different bodied AI models to run a 100m dash using Deep Reinforcement Learning. The models are rewarded for moving forward, and punished for falling over. How do you reward or punish an AI model? No clue, but if they’re anything like me, it’s leaving them alone, or talking their ear off when they’re trying to play video games. Remember: sometimes nothing is the greatest reward of all.