Is the example of a learning machine given in Martin Gardner's book a neural network?

Question

Neural networks

Martin Gardner, in Leisure Mathematics 1972, chapter 14, describes the creation of a self-learning machine for playing tic-tac-toe according to the following principle: the probability of choosing one or another step for each particular move changes after the end of the game depending on the outcome of the game. So, in fact, can such a model be considered a primitive multilayer neural network, where each probabilistic machine is a neuron?

Answer 1

Your quote is like a backpropagation neural network. But I would say no.
Read more about the minimax principle.