in this video you'll learn how to create your own ai agent using monte carlo tree search to play the game of tic-tac-toe. monte carlo tree search is likely the key algorithm that powers the o1 models for test time compute at inference time and has been used in previous ai successes such as alphago and alphazero where it's combine with deep learning .
in this video chris shows how the monte carlo tree search algorithm works and builds a simple tic-tac-toe agent, as well as comparing how monte-carlo tree search works with o1-preview and o1-mini.
github: https://github.com/chrishayuk/mcts-play