The True Story of the Chess-Playing Machine that Fooled the World
This is Tom Standage’s third pocket-sized book that delves into the history of science and engineering, and yet again he has found a subject that is not only fascinating, but which also resonates with contemporary issues. In The Victorian Internet, Standage explored the development of the telegraph, which permitted long-distance instantaneous communication for the first time. If we want to know how to cope with the Internet and the accompanying communications revolution, then we could learn something from the Victorians.
In The Neptune File, Standage explained how in the nineteenth century a new invisible planet was discovered because of its gravitational tugging, which caused Uranus to deviate from its predicted path. Today, astronomers can detect invisible planets orbiting distant stars, because the same tugging causes the stars to wobble.
Now we have The Mechanical Turk, the story of the eighteenth century automaton that convinced the world that a machine could play chess, a feat that was only truly achieved in the last decade. Standage reveals how our ancestors reacted to this first apparent example of artificial intelligence, he explains how the machine actually worked, and he brings us up to date with the terrible moment when a computer beat Gary Kasparov, the human world chess champion.
During the eighteenth century, Europe went crazy over the rise of automata, newfangled machines that seemed to mimic life. The automata emerged from the increasingly ingenious set pieces that clockmakers constructed to mark the chiming of the hour or a special feast day. These mechanical theatrical displays ranged from astronomical shows to kings and shepherds genuflecting before the Madonna and Child, presenting their gifts and retreating.
The Frenchman Jacques de Vaucanson displayed in Paris in 1737 a mechanical flute player, which could alter its breath, lips and fingers to play a tune. This was followed by a flautist that could simultaneously play the drum, and then a mechanical duck which, in Vaucanson’s own words, “drinks, eats, quacks, splashes about on the water, and digests his food like a living duck.”
But this was nothing compared to the creation of Wolfgang von Kempelen, a senior official at Viennese court of Maria Therese. In the spring of 1770 he unveiled the Turk, a life-sized figure, carved from wood, adorned with an ermine-trimmed robe and a turban. The Turk was seated behind a cabinet that was four feet long, three feet high and two and a half feet deep. On top of the cabinet was a chess set.
Kempelen would open the cabinet doors to reveal a forest of cogs, levers and clockwork machinery. Rather like a magician, he use a candle to show the audience that it was impossible to hide a human inside the automaton. He inserted a large key into the cabinet, wound up the mechanism, and the Turk was ready to play. After a pause, accompanied by clicking, ticking and whirring, the Turk moved his head, surveyed the pieces, then used his left hand to reach out and move one of the pieces. The Turk could not only move pieces, it could understand its opponents moves, think and respond accordingly.
In fact, it was a remarkably good chess player and when François-André Danican Philidor, the greatest chess player of the day, beat the Turk, he was forced to admit that no human player had fatigued him to the same extent. The Turk toured Europe and America for decades, even outliving its creator Kempelen, and wherever it went the rich and famous queued up watch and play. Ben Franklin and Napoleon took on the Turk, and Napoleon’s stepson even owned the Turk for a while.
The Turk and other automata were more than toys, because they inspired engineers and drove technology forward. For example, the computer pioneer Charles Babbage was entranced by automata as a young boy and the sight of the Turk must have fuelled his desire to build programmable machines. And the clergyman Edmund Cartwright set about building the first power loom after seeing the Turk, believing that it must be possible to be build a weaving machine if Kempelen had built one that could play chess.
Today it seems bizarre that people believed that a clockwork machine could play championship chess, but the Turk was built in the age of the enlightenment – machines were proving to be faster and stronger than humans, so why not smarter too? In the summer of 1783, the Montgolfier brothers created a flying machine, which was no more incredible than an intelligent machine.
But how did the Turk work? I am not about to spoil the ending of an intriguing book, but my favourite theory was formulated by the French magician Jean Robert-Houdin. He stated that Polish soldiers in the Russian army had rebelled and their commanding officer, Worousky, had lost both legs in the fight. He sought refuge with a Russian doctor called Osloff. During his convalescence, they played chess regularly, until Worousky became skilled in the game.
When Kempelen visited the doctor while on a trip to learn Russian, he met the Polish fugitive and came up with the idea of building the Turk. Worousky, having lost his legs, could hide inside the Turk, could be smuggled out of Russia, and could subsequently tour the world as the hidden power inside the automaton.
Standage’s book is filled with equally delightful stories, which means that the story of artificial intelligence from the Turk to today is squeezed into the final chapter. As with his other books, I feel as though I would be more satisfied with a more gradual and substantial connection between the main subject and its modern counterpart. On the other hand, if this extra material had been included, then perhaps I would complain that it intrudes on and belittles the tale of the incredible Turk.
The solution is probably a separate book that examines the role of chess in the development of artificial intelligence over the last century. Standage touches on many of the points that have arisen during this period, and each could be explored in more detail. For example, to what extent is a specialised machine truly intelligent? Even the computer that beat the world chess champion is utterly clueless in every other dimension. As the computer scientist Anatol Holt said, “A brilliant chess move while the room is filling with smoke because the house is burning down does not show intelligence.”