Monte Carlo Tree Search

I discuss creating a new board game, then training a neural network, using Alpha Zero, to play the game.