ملاحظات — كُتبت أثناء البناء

How an unbeatable Tic-Tac-Toe AI thinks: Minimax and alpha-beta, with real numbers

The AI never loses because it plays out every possible game before it moves. Here is how Minimax picks that move, and how alpha-beta pruning cuts the search from about 60,000 positions to a few thousand without changing the answer.