Can a neural network predict your next move on a game of rock, paper and scissors? That's what my last experiment is trying to answer.
It works by storing your last moves and trying to predict your next ones.
Play it here
See the code here