We study two-dimensional Ising spins, evolving through reinforcement learning using their state, action, and reward. The state of a spin is defined by whether it is in the majority or minority with its nearest neighbors. ...