The Chess Playing “Machine” That Beat Benjamin Franklin And Napoleon

In 1997, IBM’s Deep Blue supercomputer made artificial intelligence (AI) history, becoming the first machine to beat a reigning world champion –  Garry Kasparov – at the game of chess.

Though Kasparov went on to draw two other games with Deep Blue, and won a further three, it was an impressive achievement for AI. Chess, though less complex than Go, is a complicated game with a ridiculous number of possible games.

             

Defeating humans with a machine took decades of developments in computing, followed by, in the case of Google’s AlphaZero, four hours of learning the rules and learning to play.

So, in 1770, when one Wolfgang von Kempelen presented a human-defeating chess automaton to the world, perhaps people should have been a little more skeptical than they were. 

Von Kempelen first unveiled the machine to Empress Maria Theresa of Austria. It comprised of a large cabinet supposedly filled with cogs and other mechanical parts, a chess board on top, and a mechanical clockwork humanoid to shift around the chess pieces.

Before the show, von Kempelen turned the cabinet to show that it was empty but for machinery, and lifted the humanoid’s clothes to prove that there wasn’t a human or other trickery under there. He then made a big show of winding the automaton up like a clock, before it sprung into action and began to move pieces around the board.

It won its first game, impressing and confusing the court in attendance. Though there were skeptics in the crowd, and some sort of trickery suspected, Empress Maria Theresa and others were impressed and delighted with the chess “machine” (the workings of which we will get to shortly).

         

Von Kempelen toured the chess machine over the next few weeks, causing a stir around Vienna. Before long, he was asked to tour the machine around Europe, meeting and playing chess against famous figures of the time. Like an 18th Century steam punk Forrest Gump, the “Mechanical Turk” played against Napoleon Bonaparte, Charles Babbage (often hailed as the “father of computing”), and Benjamin Franklin, beating them all.

For two decades, though many suspected it was a hoax, nobody could figure out how it was performed. There were plenty, too, who believed that the “Mechanical Turk” was as mechanical as its name suggested. 

The “machine” was a decent chess player, but could be defeated, including by chess champions. It knew the rules, which would be more than impressive for a machine at the time, and played tactically. In fact, it played as well as you’d expect a decent chess player crammed inside a cabinet to perform under said conditions.

The machine was passed around owners, and continued to play players, before, in 1857, the son of its final owner revealed all in a series of articles for Chess Monthly. The machine was operated by a chess player concealed nicely within the cabinet. The chess pieces were magnetic, and corresponded to a chess board on the underneath, allowing the operator to know what move had been played. It was then up to their chess skill to defeat their opponent.

“To execute the Turk’s moves, the player engaged a pantograph, which positioned the Automaton’s mechanical arm over the playing board and operated its grasping hand,” Dan Flannigan explained in his book Legend and Lore: Jefferson Medical College. “In the center of each square of the hidden chess board, a hole had been drilled to receive the point of the pantograph. By inserting this point into the desired location, the player could execute his moves without being in visual contact with the actual playing board.”

And that’s how Benjamin Franklin and Napoleon lost at a board game, not to a machine, but to a chess player crammed inside a cabinet operating a big metal puppet.

Leave a Comment