Reinforcement Learning need to catch up with the state of the art - [ ] Alpha-go Zero (MCTS-NN) Ref: https://github.com/grimmer0125/alphago-zero-tictactoe-js