AI Notakto

Can a “computer” win a game without being programmed with a strategy? Notakto is a variation of tic-tac-toe where both players play “X” and try to AVOID making three-in-a-row. We’ll use cups to stand in for our computer, tell it a list of valid moves, and reward it for winning. Will our computer find the optimal winning strategy?

Videos:

• MENACE by Matt Parker uses a large number of volunteers and match boxes in order to make a “computer” that can play tic-tac-toe.
  • https://www.youtube.com/watch?v=R9c-_neaxeU

Popular Works:

• Reinforcement Learning 101 by Schweta Bhatt: An article which describes the basics of machine learning
  • https://towardsdatascience.com/reinforcement-learning-101-e24b50e1d292

Technical Works:

• Reinforcement Learning in Practice by Yuxi Li: A very broad discussion of reinforcement learning that cites papers where the reader may go deeper into problems that they find interesting.
  • https://arxiv.org/pdf/2202.11296.pdf