Reviewed by E.B. Heath
“ … a chess playing machine has been a holy grail since long before it was possible … I just happened to be the human holding the grail when it was finally in science’s grasp.”
Every leap forward in Artificial Intelligence technology (AI) is accompanied by alarm and despondency; dire warnings emanating from elites such as Stephen Hawking and Elon Musk really add weight to the impending peril. At the opening of the Leverhulme Centre for the Future of Intelligence at Cambridge University, Hawking commented that advancements in AI will “… be the best, or the worst thing, ever to happen to humanity”. Concerns revolve around massive unemployment, and, more dramatically, the threat to human survival if sophisticated robots make a bid to take over the world. However, Garry Kasparov, possibly the world’s best chess player of all time, has a more positive attitude towards AI and its potential for aiding human creativity and development. This is encouraging considering he suffered some, not inconsiderable, professional humiliation in 1997 at the hands of Deep Blue, IBM’s twenty million dollar chess computer. Kasparov was the world chess champion from 1986 to 2000 holding a world ranking until his retirement in 2005; he is currently a business speaker, and a global human rights activist. His latest book Deep Thinking: where machine intelligence ends and human creativity begins, is a detailed account of his loss to a machine and the implications for the future of AI in our lives.
Kasparov gives readers a brief but interesting account of the history of chess. Chess has a centuries old reputation of being a game of high intelligence and strategic thinking, and more recently has become the natural target for AI to conquer. Kasparov has been a willing partner in these aspirations, in 1985, when only twenty-two years old, he played thirty-two machines in a simultaneous exhibition. And won. In 1996 he played Deep Blue and won four to two. In a rematch in 1997 he lost by one game.
Kasparov does not like to lose and it has taken him some time to be willing to detail his experience. It turns out a chess game with a machine is not a straight forward affair, its opponents must put up with resets, crashes, reboots, and disconnects all of which can optimize the machine’s chances of a win. Another important aspect of the 1997 match was the corporation of IBM; having invested so much money losing was not an option. IBM was sponsoring the rematch consequently controlling the venue, some aspects of which were contrary to standard tournament practice viz the absence of a room where competitors consult with their seconds and the refusal to provide data of the machine’s completed games, despite the psychological gamesmanship Kasparov only lost by one game. Such was the magnitude of the global public relations around this one game victory IBM’s stock value increased by $11.4 billion in a week.
Deep Thinking gives the reader more than Kasparov’s autobiographical account of man against machine, it also analyses the strengths of AI in comparison to human attributes, and what AI can tell us about our own cognition. This is an interesting aspect of his work. AI researchers discovered that machines easily replicate high-level reasoning but low-level skills, requiring perception and mobility, require enormous computational resources. Known as the Moravic’s paradox it is theorized that difficulty reverse-engineering human skills will be proportional to the amount of time that skill has been evolving, the oldest human skills are largely unconscious and so appear to us to be effortless. Mathematics, logic and scientific reasoning are more recent human abilities and easier to replicate. Deep Blue was purpose built for chess; it had the ability to search thousands of possible moves, coming up with the best one in a matter of seconds, this ability was labeled ‘brute force’. Its tactical play was far above human ability, but not matching humans in evaluation – yet. Kasparov attributed his loss more to his mistakes under psychological pressure than to the machine’s superiority; his account of the experience is gripping.
Kasparov might feel IBM treated him poorly in his match against Deep Blue but this does not translate into a negative view of AI. He has actively contributed to chess technology, which allows a child learning chess to play at a grand master standard by simply turning on a computer. He believes the intelligent approach is to view enormous leaps in AI technology as the source of an improved humanity, allowing humans to do what they do best.
By Garry Kasparov with Mig Greengard
John Murray Publishers
ISBN: 978 1 473 6549 8