Computers, Chess and Long-Range Planning

by Mikhail Botvinnik

Mihail Moiseevich Botvinnik is an electrical engineer by profession; during World War II he headed a high-tension laboratory in the Urals and was decorated by the USSR for his accomplishments. At present, he is the head of the alternating-current machine laboratory at the Moscow Institute of Power Engineering. He is also a world-renowned chess player. He was born in 1911, and by 1935 had become a Grandmaster of Soviet chess. In 1948 he won the world chess championship and held the title until 1963 (except for a two-year break). His chess style has been characterized as deep, objective, serious, and courageous. In this book, the quality of his thinking is revealed in his study of the basic thought processes of master chess players, and his reduction of these processes to mathematical form. This formalization of thought processes is a contribution to science at three at the immediate level, it provides a basis for a computer program that seems likely to succeed in playing chess; at the middle level, game-playing programs help us to study and rationalize the processes of planning and decision-making; and, at the highest level, the study of the mind in action, as in the game of chess, leads to an understanding of human thought and of the human psyche.

Genres: Chess

89 pages, Paperback

First published July 8, 1970

Reviews

[Manny · Rating 1 out of 5]

Mikhail Botvinnik was one of the greatest chess players of all time, and famous for his logical, scientific approach to the game. He was also an expert on electrical engineering, with a doctorate to his credit. So he should be the ideal person to develop a computer chess program?

Alas, the answer appears to be no. When I first read this book, as a teenager, I assumed I was just not smart enough to get it. When I read it again, in my 20s, I knew a little about AI, and was disappointed that it still didn't make any sense. But when I read it the third time, I had to admit defeat. I suppose it's conceivable that there is some incredibly deep point I'm still missing. But then everyone else is missing it too, and, as far as I know, none of the many World Championship level chess programs that have now been developed use Botvinnik's ideas.

As Drew McDermott says in his famous checklist of common AI traps to watch out for: the fact that you can see your way though the search space doesn't mean that the machine can too.

[Liquidlasagna · Rating 1 out of 5]

Basically a World Champion with an Electrical Engineering background that just plays up the Soviet Hype about chess and computers

all promise and sizzle

[Benm · Rating 5 out of 5]

Refreshing antidote to the modern obsession with brute-force in chess computing. Still thinking about it.