Artificial Intelligence ("AI") is a branch of computer science that studies how to endow computers with capabilities of human intelligence. Playing chess or checkers against a computer is a classic example of AI usage. Speech recognition by computers is another case of AI programming. AI is, therefore, the use of programs to enable machines to perform tasks which humans perform using their intelligence. Early AI developers avoided dealing with most aspect of human psychology models, but in recent years the development of connectionism, which is based on theories of how the brain works, paved the way to the introduction of human psychology concepts into AI programming. In connectionism, complex functions, including learning, involve the transmission of information along pathways formed among large arrays of simple elements. Ever since the old days of punch-card computers, people have been fascinated with chess-playing and checkers-playing programs, and no wonder. If humans can create a computer that can outsmart people in these obviously strategic, thinking-based games, then it would be a major milestone on the road to the Intelligent or Smart Computer.
Chess and Checkers are games of perfect information. This means that unlike poker or backgammon, all information specific to these games is known to both game players, and there is neither secrecy nor chance. Both players can see everything about the position on the chess or checkers boards, and they know each other's moves, for the moves are not played at the same time but rather in order. Thus, chess and checkers are purely intellectual games, perfect environments for testing Artificial Intelligence techniques. Years ago, many researchers concluded that the computer's greatest asset was its computational speed. They determined that computers would utilize a variant of the brute-force routine, an algorithm that relies almost purely on how fast the computer can process the board. Chess and Checkers have a small board: 64 squares. Chess has up to 32 pieces to fill them and Checkers has up to 24 pieces to fill them. Even though there are an astounding number of legal chess or checkers positions, the researchers relied upon the fact that computers could process more exactly and more quickly than people, and thus "think" faster than people while playing the game.
In those days, of course, computers had a ridiculously small amount of memory of 300 K of memory capacity and processing power, so chess and checkers programs were, at best, rudimental. In recent years computers became much faster, allowing chess and checkers programs to become much more intelligent. With the combined research of many brilliant artificial intelligence programmers, the minimax game tree evolved to become the most major component of the chess program. Almost all chess and checkers programs use the minmax tree in one form or another. It is by far the most successful and most popular of all the AI techniques in the history of board game programming.
In May 1997, IBM's Deep Blue Supercomputer played a fascinating match with the reigning World Chess Champion, Garry Kasparov. Deep Blue II defeated Kasparov.
There are many AI techniques that can be implemented and tried in a Chess and Checkers programs, such as neural networks, genetic algorithms, and collaborative computing. Chess and Checkers programming are done in controlled environments in which the computer is presented with a situation and a target goal, and the computer must find possibilities and make decisions to achieve that goal.