The hotly debated topic of artificial intelligence seems to stir up either a pie-eyed optimism or an almost religious impulse to doomsay. Garry Kasparov has faced far-reaching personal consequences as a result of AI development. Now, the chess legend has written a new book, Deep Thinking, about his experiences battling machine intelligence on the board and what he, and all of us, can learn from them.
Much has been said and much written about the rise of artificial intelligence. To some, like the futurist Ray Kurzweil, it presents a glittering new horizon, one that reveals a universe of technological potential as yet undreamed-of. For others, like the physicist Stephen Hawking, it summons up fears of apocalyptic horror: the mind, after all, is what makes us unique. Whither humanity once it has been recreated by technology?
Garry Kasparov — for full disclosure’s sake, we will note here that he sits on The Octavian Report’s editorial board — is considered by many to be the greatest player in the history of chess. He is also one of the few people to have experienced in a direct, personal, and antagonistic way the exponential increases in power of artificial intelligence since the end of the Second World War. His new book, Deep Thinking: Where Machine Intelligence Ends and Human Creativity Begins (PublicAffairs, 285 pp., $28), recounts and reflects in his inimitable, engaging prose on his long involvement with, and eventual battles against, the ever-evolving suite of chess AI programs that proliferated during the Cold War and afterwards. Deep Thinking also serves as an introductory history of the theoretical and mechanical origins of artificial intelligence, which begin with the visionary musings of Alan Turing and Claude Shannon and continues today in the secretive special projects divisions of Alphabet (the corporation formerly known as Google) and high-level university research labs throughout the world. The book’s third, and perhaps most intriguing element, is an extended meditation on the nature of mind, human or machine: how they differ, where each is strongest, the difficult and painful results of their competition, and how they might work together. Despite Kasparov’s eventual loss of a match to IBM’s chess-playing computer Deep Blue, he remains far more optimistic about what the rise of machine thinking means for humanity than one might expect.
Deep Thinking’s main narrative focuses on the development of chess AI and its interactions with human players. The drive to develop a computer capable of playing chess begins, as noted, with Turing, who was able to build a fairly decent chess program using only a set of algorithms written down on slips of paper. Efforts at increasing the strength of AI players really took root in the early years of the Cold War, along similar political lines: Russians and Americans competing over a technological marvel, with the Russians enjoying an early and long-lasting lead. The first major human-computer showdown took place in 1963 between the iconoclastic Russian Grandmaster David Bronstein and the Soviet chess computer called M-20. Bronstein beat M-20 handily. But as processing power became cheaper and human understanding of algorithms became deeper and deeper, chess AI grew ever more powerful. The Scottish International Master David Levy, Kasparov recounts, made a famous bet in 1968 than in a decade no computer would be sophisticated enough to beat him — and he was proven right, despite the fact that the computer player eventually proved to be far tougher to best than he expected. Levy experienced, years later, the first loss of a chess game by a human to a computer, and would go on to be defeated in a match by an ancestor of the machine that Kasparov himself would eventually take on in two historic matches. The first resulted in a victory for the Russian legend and the second in a shocking defeat — about which, Kasparov quite candidly admits, he still has some hard feelings. “I hate to lose,” he says.
“Despite Kasparov’s eventual loss of a match to IBM’s Deep Blue, he remains far more optimistic about what the rise of machine thinking means for humanity than one might expect.”
The story of those two matches forms the emotional core of the book, and Kasparov’s gifts as a writer and raconteur are such that even non-players will be swept up in the intense intellectual and psychological drama that defines high-level chess. Kasparov describes himself as the best-prepared player in the history of the game (referring to his habit of intense training and research into openings, past matches, and other recondite areas of study prior to play) going head to head with a machine capable of perfect recall, millions of brute-force calculations per second, and no human frailty. But Kasparov points out a paradox first codified by Hans Moravec: that high-level reasoning requires much less processing power than basic sensory/motor perception. He also notes that human strengths and machine weaknesses complement each other. Deep Blue might have had immensely superior calculating powers, but Kasparov, as a human being, possessed gifts of perception simply unavailable to machines. Kasparov eventually lost one match to Deep Blue. He was deprived of another chance to play the machine by the polite refusal of IBM’s management — something he notes was a very canny piece of marketing by then-CEO Lou Gerstner. But in the course of their struggle “a real battle, rich game of chess” had taken place, says Kasparov. He means, of course, that Deep Blue was playing on a human or quasi-human level.
Now, the subject of whether this victory was rightfully earned has been a source of debate. Kasparov puts that to bed in a chapter near the book’s end: “After twenty years of soul-searching, revelations, and analysis,” he writes, his answer to the question of Deep Blue cheating is a firm “no.” This “no” makes the book’s conclusion all the more important. In Deep Thinking’s final two chapters, Kasparov broadens his theme and his focus on the past and looks toward the future — as any great chess player must, of course, do. He starts off with some bold statements. “Do not confuse nostalgia with the loss of our humanity,” he says apropos of tech journalist Cory Doctorow’s famous essay “My Blog, My Outboard Brain.” Technology, Kasparov argues, has liberated us to the point that those living in developed societies enjoy previously unimaginable amounts of spare time — and it is on the cusp of providing humanity with massive assists in cognitive ability to use that time. He explains this through the lens of the now almost-universal acceptance of analytical engines in chess around the world: not as opponents but as tools for human to become better and stronger players. A lesson, he clearly believes, that is adaptable to many other fields of human achievement. Provided, of course, that we can set aside our existential anxieties about AI and embrace the huge potential they can offer.
He ends with an evocation of that potential as stirring as it is clear-eyed:
“This is not a choice between utopia and dystopia. It is not a matter of us versus anything else. We will need every bit of our ambition in order to stay ahead of our technology. We are fantastic at teaching our machines how to do our tasks, and we will only get better at it . . . We need new frontiers and the will to explore them. Our technology excels at removing difficulty and uncertainty from our lives, and so we must seek out ever more difficult and uncertain challenges.”
Whatever risks and social changes AI brings about in the next decades, it’s fortunate that we will have at least one futurist who is neither a digital-addled Pollyanna or a gloomy Luddite. And who knows? Perhaps Mr. Kasparov will get the rematch he’s been hoping for.