In just 4 hours Deepmind AI trained it self to play chess, without the use of any chess play libraries and was able to defeat the world champion chess-playing program, Stockfish. After 100 games It had not lost a game and won 24.
One of the key advances here is that the new AI program, named AlphaZero, wasn’t specifically designed to play any of these games. In each case, it was given some basic rules (like how knights move in chess, and so on) but was programmed with no other strategies or tactics. It simply got better by playing itself over and over again at an accelerated pace — a method of training AI known as “reinforcement learning.”