I recently read Brian Christian’s book The Most Human Human, in which Christian uses his experience as a “confederate” in the 2009 Loebner Prize competition to see what machine intelligence can teach us about our own human intelligence.

In one particular section, Christian examines the famous 1997 chess match between Grand Master Gary Kasparov and IBM’s Deep Blue, to understand how human thought is conceptualized.

Chess has long been associated with mental faculty. Benjamin Franklin, in his article “The Morals of Chess” (1750), wrote:

“The Game of Chess is not merely an idle amusement; several very valuable qualities of the mind, useful in the course of human life, are to be acquired and strengthened by it” 

at the time, this idea of chess, and its mathematical nature, made it the most logical thing for A.I. researchers to try and emulate to be thought of as close to human (like Watson did with Jeopardy). However After the match, Christian notes that that their was a tendency not to think of Deep Blue as reaching human intelligence, but to start thinking of chess as trivial. Christian sites Douglas Hofstadter as saying:

“Once some mental function is programmed, people soon cease to consider it as an essential ingredient of ‘real thinking.”

This tendency occurs partly from the demystifying of an action (a chess computer program access and lays out millions of possible steps between each action) and partly out of defence of the ‘special’ nature that we give to human intelligence. This begs the question, “What is human intelligence?”, What is thinking? and Is there an aspect that cannot be conceivably duplicated by machines in the future? While some of these don’t have knowable answer, Christian finds something in his Loebner experience; computers currently have difficulty with improvization and spontaneity. 

In examining the Deep Blue match Christian relates that computers and chess masters alike take part in a form of memorization, where past games, strategies, moves, and matches are studied and put into long-term storage. This memorization creates the ‘book’ of chess, the standard that all serious player know.One can play ‘by the book’ or venture off into the unknown. Kasparov, as player, is known for going “Out of Book.” Going out of book suggests such a mastery of the game that one can reshape it. At a certain level, it is what one plays the game for, to play a match that is closer to chaos and therefore stimulating. In the games that Kasparov won the match went “out of book.”In the final game of the 1997 match Kasparov made a blunder and felt that that game didn’t count because it was too “by the book.”

There are some thing I like about going “out of book.”  When more and more activities are being combined with the web, finding that “out of book” moment is ever more valuable (all of art is based on it). However improvization and spontaneity seem like a temporary human advantage. I can conceive of A.I. technology that performs a human version of improvised and irrational conversation and I wonder if their is a time where it will be hard to find activities that don’t feel ‘programed.’ I am eager to go back to the brain and learn as much about its functioning as possible and find its long term advantages.

note: One of Kasparov’s solutions was to merge with machines and play Advanced Chess