Chess and Computers

Edward Winter




In 1989 Garry Kasparov offered some comments on chess computers in an interview with Thierry Paunin on pages 4-5 of issue 55 of Jeux & Stratégie (our translation from the French):

‘Question: ... Two top grandmasters have gone down to chess computers: Portisch against “Leonardo” and Larsen against “Deep Thought”. It is well known that you have strong views on this subject. Will a computer be world champion, one day ...?

Kasparov: Ridiculous! A machine will always remain a machine, that is to say a tool to help the player work and prepare. Never shall I be beaten by a machine! Never will a program be invented which surpasses human intelligence. And when I say intelligence, I also mean intuition and imagination. Can you see a machine writing a novel or poetry? Better still, can you imagine a machine conducting this interview instead of you? With me replying to its questions?’

(1802)



A nomination made by Leonard Barden in 1967 for the best computer move:

dia

This position, from a game between the Moscow Institute and Stanford University, appeared on page 187 of the June 1967 Chess Review. Barden’s view on White’s next move, 15 Rxh7, was quoted: ‘Undoubtedly the most brilliant move ever made by a computer’.

The Chess Review item, by J.S. Battell, ended:

‘We don’t know what brilliant moves may have been made by computers previously; but, in this game, the Moscow machine looms like a Morphy against a Duke of Brunswick and a Count Isouard. We do expect, however, that the future World Champion who takes on a computer will have a referee so stationed as to ensure against there being a Schlumberger behind the machine’s moves!’

(2234)



A news report by Harry Golombek on pages 8-9 of the January 1952 BCM:

‘My more mechanically minded readers will be interested to learn that a portable electronic brain, weighing a mere 500 pounds and costing only 80,000 dollars, has been developed by the Computer Research Corporation of Hawthorne, California. One of its designers, Richard Sprague, claims it can play unbeatable chess. Donald H. Jacobs, president of the Jacobs Instrument Company of Bethesda, Maryland, and himself developer of a 140-pound mechanical brain, proved sceptical and challenged the CRC-102 (the euphonious and imaginative name of the electronic midget brain) to a best of 20 games match, offering to bet 1,000 dollars on his ability to defeat the baby electronic brain without consulting his own mechanical brain for assistance. The CRC-102 declined the match hastily on the grounds that the “urgency for this machine in the defence effort makes such a tournament untimely”. Clearly history is engaged in its habitual process of repeating itself and we are faced once again with the Morphy-Staunton incident.’

(2646)



The following is Fred Reinfeld’s introduction to Tartakower v L. Steiner, Warsaw, 1935, on page 116 of Relax with Chess (New York, 1948):

‘When Professor Weiner [sic – Wiener] of the Massachusetts Institute of Technology invented a calculating machine which requires only one ten-thousandth of a second for the most complicated computations, he was quoted as saying, “I defy you to describe a capacity of the human brain which I cannot duplicate with electronic devices”.

Up to the time these lines were written, the Professor had apparently not yet perfected an electronic device capable of making such chess moves as Tartakower’s 20th in the following game. The day may yet come, however, when we shall see such books as ‘Robot’s 1000 Best Games’, or when chess tournaments will have to be postponed because of a steel shortage.’

Today’s computers find 20 Nxd8 instantaneously. As regards his 19th move, i.e. the sacrifice Nxf7, Tartakower wrote on page 61 of his second Best Games volume: ‘The art of chess is simple: you play Nf3-e5 and then, sooner or later, Nxf7 is decisive.’

(3177)



Jonathan Hinton (East Horsley, England) draws attention to the following passage by P.H. Clarke in his ‘From the USSR’ column on pages 117-118 of the April 1963 BCM:

‘Chess in 2000

Will the world champion in 2000 be a man or a machine? This was the subject for discussion on Moscow television recently by grandmasters Smyslov, Bronstein and Averbakh. As might be expected, no agreement was reached. Averbakh held that in about 35 years’ time scientists will be able to design a machine perfect enough to compete with masters and even grandmasters and that it will have many advantages over man. Smyslov was of a different opinion. He compared chess with music, asserting that just as a mechanical composer could not rival human fantasy, so a machine could not play better chess than a man.

Bronstein took a middle line, surmizing that by the year 2000 there would be separate championships for men and machines. I wonder who will be right.’

(3846)

Stuart Rachels (Tuscaloosa, AL, USA) compares the above with a comment of Smyslov’s, dated 30 April 2004, which is quoted on page 122 of The World Champions I Knew by Genna Sosonko (Alkmaar, 2013):

‘I’m working on a book – my 60 best games, I was looking at my game with Savon recently. And I found so many mistakes with the computer, just one mistake after another. And I considered that game one of my best … Yes, the computer can outdo anyone now.’

(8332)



From page 345 of The Personality of Chess by I.A. Horowitz and P.L. Rothenberg (New York, 1963):

‘That a richly endowed robot will one day be able to play a highly skillful game of chess leaves no room for doubt. On the other hand, in the absence of a fantastic superspeed electronic brain, the chess championship of the world is likely to be retained by humans for centuries to come.’

(4126)



From Nathan Bauman (Seoul, South Korea):

‘Regarding the first chess computer, a king and rook versus king problem-solving machine invented by Leonardo Torres y Quevedo (1852-1936), I would draw your attention to a recent article I have written, “A New Photograph of ‘El jugador ajedrecista’, the World’s First Chess Computer”.

I wonder whether any other pictures or clips can be found. There is very little visual material available on the Internet, and what does exist seems to have been conflated with a machine invented by Torres y Quevedo some years later.’

Any references that readers can provide will be appreciated. In the meantime, we note that some illustrations of Torres’ invention (a photograph and four diagrams) appeared on pages 227-229 of that much-neglected book Chess: Man vs Machine by B. Ewart (London, 1980).

Although Ewart noted that Torres’ machine (‘the first purely mechanical chess player’) was introduced to the public at the Sorbonne in Paris in November 1915, a number of chess computer books give the date of its invention as 1890. See, for example, page 11 of Computer Chess by L. Pachman and V.I. Kühnmund (London, 1986) and page 20 of I giocatori artificiali by P. Ciancarini (Mursia, 1992).

Concerning the subsequent generation (in terms of both computers and the Torres family), below is an item from page 94 of the February 1951 CHESS:

tartakower

(4470)

From Christian Sánchez (Rosario, Argentina):

‘On the Internet I have found two reliable Spanish sources for information about the chess machines. One is a webpage of the Torres Quevedo Museum, where the two prototypes are preserved; there are pictures of both Ajedrecista machines. The other is an article by González Redondo; it has some illustrations of the first Ajedrecista and allows a chronology of the inventions to be drawn up:

1912: Leonardo Torres Quevedo began construction of the first Ajedrecista.
1914: The first Ajedrecista was exhibited in Paris.
1920: The second Ajedrecista was built by Gonzalo Torres Quevedo, under his father’s direction.
1922: The second Ajedrecista was exhibited in Paris.’

Calle Erlandsson (Lund, Sweden) points out that in 1955 Spain issued a postage stamp depicting the inventor:

torres

We note that the stamp was one of the illustrations in the section on the Spaniard on page 125 of Persönlichkeiten und das Schachspiel by Beat Rüegsegger (Huttwil, 2000).

ruegsegger

(4482)

Leonard Barden (London) comments that Torres’ invention was placed on view during the 1960 Olympiad in Leipzig. The photograph below appeared on page 253 of XIV. Schach-Olympiade Leipzig 1960, published by Sportverlag Berlin:

torres

See also plate 1 in The Machine Plays Chess? by Alex G. Bell (Oxford, 1978).

(4495)

An unresolved matter from C.N. 4470 is why a number of computer chess books affirm that Leonardo Torres y Quevedo invented his chess computer as early as 1890, rather than in 1912-14.

The earliest claim about 1890 of which we are aware is on page 286 of Faster Than Thought edited by B.V. Bowden (London, 1953):

‘In about 1890 Signor Torres Quevedo made a simple machine – a real machine this time – which with a rook and king can checkmate an opponent with a single king.’

The Italianization does not inspire confidence, and no source for the information was offered.

Page 9 of The Machine Plays Chess? by Alex G. Bell (Oxford, 1978) stated:

‘Torres apparently built a prototype machine about 1890 which went through a number of sporadic refinements. His final version made its most publicized performance at the Paris World Fair in 1914 but it, understandably, sank into oblivion on the outbreak of World War I. So much for the discrepancy in dates.’

Can a reader move the matter on from the ‘apparently’ stage to some hard facts?

(4525)

It may be wondered when 1890 started to be mentioned in chess books as the year in which Leonardo Torres y Quevedo invented his chess computer. A relatively early instance is page 13 of Chess and Computers by David Levy (London, 1976), whereas page 3 of the same writer’s Computer Chess Compendium (London, 1988) gave only 1914.

(4547)

From opposite page 96 of Schach – mehr als ein Spiel by Herbert R. Grätz (Leipzig, 1964):

computer



‘This week, in Chicago, a robot, controlled by a photo-electric cell, will take on all comers at chess. The presumptuous creation is even offering $150 to anyone who can outwit his moves engendered by the electric-eye principle. O ye spirits of the great and ye living masters. Evans and Steinitz, Euwe and Alekhine. Gang up mercilessly on this soulless monster, split the profits and, most of all, re-establish the dignity of the human mind.’

That call to arms appeared in the Chicago Daily News and was cited on page 109 of the September-October 1938 American Chess Bulletin. What more is known about the robot?

(4581)



From page 357 of the December 1949 Chess Review:

‘The claim is made that it is now possible to solve two-move chess problems automatically by means of an electronic calculating machine. Even if this device portends the invention of a monster that can play a whole game, the world of human chess, it is safe to predict, will stick resolutely to its time-honored methods of enjoying the royal game. It is fascinating, however, to speculate on whether a machine could ever defeat a Botvinnik and outstrip Ambrose Bierce’s fantastic thriller Moxon’s Master.’

(7536)



From Mark McCullagh (Belfast, Northern Ireland):

‘Which game is the first recorded loss by a FIDE-titled player to a computer at any time control?’

(5496)

Owen J. Clarkin (Ottawa, Canada) writes:

‘Page 60 of More Chess and Computers by D. Levy and M. Newborn (Potomac, 1980) states that the program Chess 4.6 defeated Hans Berliner, Lawrence Day, Robert Hübner, David Levy, Michael Stean and Zvonko Vranešić in blitz games in 1977. The earliest game-score given is a March 1977 victory over Berliner.’

(5526)

C.N. 1821 (see page 81 of Chess Explorations) quoted a claim on page 60 of John Curdo’s Chess Career. Forty Years at the Top (Coraopolis, 1988):

‘I believe to this day that I am the first master to engage a computer program in rated tournament play.’

The game, against Mac Hack, was played at the Greater Boston Open on 9 October 1971 and is available in databases.

(5634)



Caleb Wright (Ohauiti, Tauranga, New Zealand) asks for solid information about reports that the Soviet computer KAISSA helped David Bronstein in an adjourned game in 1975.

Firstly, we quote a passage from page 128 of How Computers Play Chess by D. Levy and M. Newborn (New York, 1991):

‘The first occasion on which a program’s ability to play certain endgames perfectly was useful to human players was at a Soviet tournament in 1975. There, in Vilnius, Grandmaster David Bronstein was able to use a database created by the KAISSA team to help him analyze an adjourned game, which he subsequently won. This was the ending of queen and knight’s pawn (g-pawn or b-pawn) against queen, the most difficult of all queen and pawn endings.’

Such an ending, we note, occurred in Bronstein’s game in Vilnius against Karen Grigorian. More details, from primary sources, of the computer’s involvement will be appreciated.

For other information on KAISSA, see chapter six of Chess in the Eighties by D. Bronstein and G. Smolyan (Oxford, 1982).

(6150)



From page 210 of Every Great Chess Player Was Once A Beginner by Brian Byfield and Alan Orpin (Secaucus, 1974):

chess

(6363)



An interview with Botvinnik on computers in Soviet Weekly was reproduced on pages 194-195 of CHESS, 25 February 1961.

A cautious reply from Botvinnik to an interview quoted on page 120 of CHESS, January 1962:

‘In how many years do you think chess by electronic computers will become a serious factor in the game?’

‘I believe the time when an electronic machine will begin to play chess is not far off.’



Page 8 of CHESS, October 1968 included the following remark by Botvinnik when addressing 400 people in Vladimir in late July:

‘I forecast an unprecedented period of popularity for the game. When an electronic machine has started playing chess and played it successfully this will be such a momentous event that every schoolboy will want to know about it. In world history, it will perhaps fall not far short in importance of the discovery of fire.

The young will have to study not only computer technique and programming but also chess itself. And then when a hundred times more young people study chess, when many of them devote their lives to it, then we shall have a real chance of getting a new generation of Tals and Spasskys.’



Early cartoons depicting computers/machines and chess are always welcome. From page 65 of the December 1946 CHESS:

cartoon mountbatten

(8464)



‘Is this the beginning of the end for chess, or the end of the beginning?’

From page 38 of CHESS, November 1949:

computers

(8582)



From page 141 of A History of Chess by Jerzy Giżycki (London, 1972):

computer



lasker shannon

The Adventure of Chess by Edward Lasker (New York, 1950), opposite page 201

shannon lasker

Chess Review, January 1951, page 13.

(10324)

An article ‘Experiments in Chess on Electronic Computing Machines’ by P. Stein and S. Ulam on pages 13-16 of the January 1957 Chess Review was introduced by Edward Lasker and ended with three games of ‘6x6 chess’. For example:

6x6 chess

The three games, together with a summary of the rules, were published in the entry on Los Alamos Chess on pages 175-176 of The Encyclopedia of Chess Variants by D.B. Pritchard (Godalming, 1994).

(10325)

On the front cover of the July 1959 Chess Review:

roberto

From page 199 of the same issue:

roberto

(10377)



Latest update: 28 March 2017.

To the Chess Notes main page.

To the Archives for other feature articles.



Copyright: Edward Winter. All rights reserved.