A reinforcement learning agent that plays tic-tac-toe
This was inspired by the first bit of Reinforcement Learning: An Introduction by Sutton and Barto. Something about the tic-tac-toe thought experiment seemed almost too simple, so I figured I'd try it :)