Artificial Intelligence, or AI for short, is a rapidly growing area of technology that has the potential to change many aspects of our lives. It assists us with problems that would normally require human intelligence, such as information retrieval, learning, decision-making and, most importantly, language processing. The impact of AI is already significant. Translations are fluent, navigation in road traffic is becoming more and more efficient and with a simple question to an AI system we receive extensive answers with a high degree of reliability - and on all topics.
The game of chess is also significantly influenced by artificial intelligence today, whether for learning or playing chess. Our understanding of chess has been revolutionised with the advances in algorithms and heuristics and has since shaped the analysis of games but also the predictions of moves.
But the use of computers and AI in chess has a longer history. In the following, I would like to give you an overview of the events - from the first "chess computer" to today's technologies.
The Early Stages of Artificial Intelligence in Chess
Even though there was no knowledge of artificial intelligence at the time, it was the first chess robot that gained fame as an automaton. In 1769, Wolfang von Kempelen constructed the Mechanical Turk. A supposed robot that was able to play above average and defeat most players. Even though Kempelen always described his invention as a mechanical trick, it amazed many people and became internationally famous. Later it turned out that humans had operated the robot.
The Chess Turk by Wolfgang von Kempelen
(copper engraving by Racknitz - 1789)
Even if the Mechanical Turk did not have artificial intelligence, it became apparent that the idea that a machine could be superior to humans fascinated people early on. Efforts to develop artificial intelligence became more concrete when the first computer scientists developed algorithms that could at least rudimentarily model human play.
Konrad Zuse, for example, wrote the first chess program in the mid-1940s. Alan Turing and David Champernowne designed the "Turochamp" in 1948, the first algorithm that assigned a value to chess pieces and was supposed to determine the best possible chess move based on all possible positions. Even modern chess programs today still work according to this principle. The publication of the first chess algorithms was followed by others. In 1949 Claude Shannon developed the Minimax algorithm, in 1961 Dietrich Prinz wrote the first computer program that could solve chess problems, and in the mid-1950s John von Neumann implemented chess in a game-theoretical concept based on a two-person zero-sum game.
In the 1960s, it slowly became apparent that computers would be able to compete against humans in the future. In 1967, the "Mac Hack" developed by Richard Greenblatt became the first computer program to compete in a tournament, achieving a rating of 1239. In order to make the development of chess programs accessible to all chess players, Peter Jennings distributed the First Chess Program for Home Computers from 1976 and called it "Microchess". Although the special rules of castling, en passant and pawn promotion were not implemented, it sold with great success. The chess program "Belle" by the two developers Ken Thompson and Joe Condon finally achieved the breakthrough when it participated in the 1983 American Chess Championship and achieved a rating of 2363 and the title of National Champion. "Belle" was the leading chess software of the 1980s and is considered a direct precursor of artificial intelligence as we know it today.
Modern Artificial Intelligences in Chess
A turning point in the development of artificial intelligence in chess was the games of the computer Deep Blue against the then world champion Garry Kasparov. Deep Blue won under tournament conditions and gained worldwide attention as an artificial intelligence that defeated humans in chess.
The development of Deep Blue began back in the 1980s, when researchers at IBM started working on a computer that could play chess. The first version of the programme was called "ChipTest" and was capable of playing a simplified version of chess. Over time, the researchers optimised the programme and it became more sophisticated and powerful.
Garry Kasparov - World Chess Champion from 1985 to 2000
(Copyright 2007, S.M.S.I., Inc. - Owen Williams, The Kasparov Agency.,
CC BY-SA 3.0, via Wikimedia Commons)
In 1996, Deep Blue played its first game against Kasparov, but the computer defeated him. Kasparov won three games, obtained two draws and lost only once. The IBM team learned from their mistakes and set to work to improve the programme further. They added more powerful hardware, improved the search algorithms and made further optimisations to the software.
In 1997, Deep Blue faced Kasparov again in a rematch. The match gained a lot of media attention and was eagerly awaited worldwide. At that time, Deep Blue was able to calculate 200 million positions per second and delivered a much better game than the year before. Kasparov won the first game, but surprisingly gave up in the second in an apparent draw position. The third game ended in a draw and Deep Blue won the fourth. After further draws in the fifth and sixth games, the winner was finally decided. Deep Blue defeated World Champion Garry Kasparov by 3.5 points to 2.5.
Deep Blue, a computer similar to this one defeated
world chess champion Garry Kasparov in May 1997.
(Copyright: James the photographer,
CC BY 2.0, via Wikimedia Commons)
Deep Blue's victory over Kasparov was a significant moment in the history of artificial intelligence and chess history. Computers were now able to defeat the best human players in the world and provide a completely new perspective on chess. The power of artificial intelligence was even predicted to revolutionise the use of computers in other human activities - a claim that turned out to be true and that today has a major impact on our lives.
Since the match between Deep Blue and Kasparov, artificial intelligence has developed rapidly in the field of chess. Particularly through the internet and machine learning, there are now many intelligences that can be played against but also used to analyse one's own playing style.
Known Artificial Intelligences in Chess:
Stockfish
Stockfish is a free and open-source chess engine known for its strength and accuracy. It is widely regarded as one of the strongest chess engines in the world and is used by many players and analysts to study games and prepare for tournaments.
AlphaZero
AlphaZero is an AI developed by DeepMind, the company that also developed the famous Go playing AI AlphaGo. AlphaZero is unique in that it learns the game of chess from scratch, without any prior knowledge or human input. It has achieved remarkable results, defeating some of the strongest chess engines in the world.
Leela Chess Zero
Leela Chess Zero is another AI that uses "deep learning" to improve its game. It is trained with a process called "reinforcement learning", where it plays games against itself and learns from its mistakes. She has achieved impressive results and won several computer chess championships.
Houdini
Houdini is a commercial chess engine known for its tactical and aggressive play. It has won several computer chess championships and is used by many top players and coaches to analyse games and study openings.
Komodo
Komodo is another powerful chess engine, which is widely used in professional tournaments. It has won several computer chess championships and is known for its strategic and positional play.
The Future of Artificial Intelligence in Chess
The development of chess programs and artificial intelligence has significantly shaped and fundamentally changed the game of chess and the learning of chess over the last 30 years. Today, players can learn the rules of chess in a very short time with digital exercises, play together with other players around the world in a matter of seconds and have their playing style analysed by powerful algorithms. With the help of artificial intelligence, many chess players of tomorrow will be able to defeat the world chess champions of yesterday. But the modern intelligences themselves, will increasingly seem invincible to humans.
I am glad that I could give you a short overview of the chess history of artificial intelligence. If you have any further questions, please feel free to write to me via my contact form. Besides computers, playing chess on a real chess board is still popular. If you are interested in chess pieces or chessboards in tournament format, please have a look at my assortment.
I wish you a lot of fun with the game, much success and rapid progress in your learning.
See you soon.
Stefan