I am dreadful at chess. I remember in 1983 trying to beat my Atari 800XL at chess, but even on level one, I could never win. In 1983 the predetermined intelligence of a computer program had me defeated. Fourteen years later, IBM’s Deep Blue chess computer beat the intellectual titan of the chess world- Garry Kasparov. (Harari 2018). The natural intelligence of Garry Kasparov had been defeated by the artificial intelligence of Deep Blue.
The defeat of Kasparov by Deep Blue changed the world of chess forever; chess players began to learn from the strategies adopted by advanced chess computer algorithms. Human chess players improved their chess-playing abilities by learning from their silicon-based opponents.
In 2016 the most powerful chess computer program on the planet was Stockfish. Twenty years after Garry Kasparov’s defeat by the transistors of Deep Blue, Stockfish was defeated in 2017 by a new type of machine in AlphaZero.(Silver et al. )
Stockfish had access to the complete knowledge of aeons of human chess experience and decades of computing chess practice. But AlphaZero had a new type of intelligence, it possessed machine learning. Before its tournament with Stockfish, AlphaZero taught itself chess in a few hours by playing against itself, with only having been supplied with the rules of chess. In the 100-game contest, AlphaZero did not lose a single game and defeated Stockfish in 28 and drew 72 games. (Silver et al. ) In just 4 hours, AlphaZero was able to learn and surpass Stockfish’s knowledge of the entire human/computer experience of chess. AlphaZero had gone from a complete lack of knowledge to an ingenious command of chess in four hours without any anthropological chaperone. (Harari 2018)
The world of chess gives an insight into the power of artificial intelligence and machine learning. Natural human intelligence in chess was defeated 22 years ago by artificial intelligence, and today advanced machine learning techniques are giving artificial systems some quite remarkable capabilities.
Harari, Y.N., 2018. 21 Lessons for the 21st Century. Random House.
Silver, D., Hubert, T., Schrittwieser, J., Antonoglou, I., Lai, M., Guez, A., Lanctot, M., Sifre, L., Kumaran, D. and Graepel, T. A general reinforcement learning algorithm that masters chess, shogi and go through self-play.