The Mystery of Mastery

×

Expand

Mechanical Turk

While helping me with research for a forthcoming talk on robotics, my unpaid research team (wife Paulette and daughter Kim) came across the recent article “Chess and 18^th Century artificial intelligence” by Adam Gopnik, an American commentator and a staff-writer for The New Yorker.  The article is about the Mechanical Turk, a chess-playing machine constructed in the late 18^th century.   

The Turk, as the machine is commonly known, seemed able to play chess against a human opponent.  The machine gave the appearance of possessing perception and reasoning skills, in other words, what we now call ‘artificial intelligence’.   It was able to win games against many challengers including, purportedly, famous personalities such as Benjamin Franklin and Napoleon Bonaparte.

Unfortunately, some 50 years after its first demonstration, the Turk was exposed as a hoax.  Hidden in its cabinet, together with the gears, cogs, pulleys and levers, was a skilled human chess player who operated the machine.  The Turk was not the automaton that it had been portrayed — and believed — to be.

Mr Gopnik drew a number of observations from the story of the Turk.  

His first observation relates to what he calls “an odd, haunting hole in human reasoning.” He argues that common sense should have told the people who watched the Turk that, for it to have been a genuine chess-playing automaton, “it would have had to have been the latest in a long sequence of such machines. For there to be a mechanical Turk who played chess, there would have had to have been, 10 years before, a mechanical Greek who played checkers.”

One might dispute this comment as it seems to disregard the possibility of technological disruption.  Such a disruption would indeed allow the creation of ‘mechanical Turks’ without the prior invention of ‘mechanical Greeks’. However, it would be difficult not to agree with Gopnik’s observation that, by and large, people were fooled because they were looking “for the beautiful and elegant solution to a problem, even when the cynical and ugly one is right.”

Here, the “beautiful and elegant solution” is the intelligent chess-playing automaton and the “cynical and ugly solution” is the human player concealed inside the machine cabinet.  Looking for beautiful and elegant — perhaps more precisely termed ‘sophisticated’ — answers to problems, when simpler (cynical and ugly) alternatives exist, is a common mistake which I sometimes struggle to teach novice researchers to avoid.

Perhaps Gopnik’s most interesting remarks concern the so-called ‘asymmetry of mastery’ — the mystery of mastery.  People tend to overrate masters while underrating mastery.  In the 18^th century, they were unable to imagine that there could ever be a chess master small enough to hide inside the Turk.  However, it was chess mastery that was needed of the human operator in the Turk and highly competent but not champion players were readily available who could fit in the Turk.  The hoax required someone who was very good at chess rather than someone uniquely good at it.

Do universities need masters or mastery?  As academic life is about truth and genuine excellence, I would contend that those of us who are in academia should strive to become masters.  However, we should bear in mind that many of us will have to be satisfied with being just very good at our subjects, because only few could reach the status of unique virtuosos.  Then, as we are unlikely to be able to claim uniqueness, in our efforts to master a subject, it is sobering to remember that, once we have gained that mastery, we are almost certain to find someone else who is better than us.