The devil & Deep Blue
Chess grand master Garry Kasparov believes true machine intelligence is about to render most careers obsolete.
Garry Kasparov has a distinct take on history. It is, he tells us, the story of humans automating their own labour, replacing themselves with machines – slowly at first but with increasing rapidity since the industrial revolution. Until now, the labour has been physical, but we’re on the threshold of a machine-intelligence revolution in which most of our cognitive tasks will be performed by computers.
Kasparov is one of the greatest chess players of all time. He was ranked No 1 in the world for almost 20 years, an astounding achievement that seems a little less impressive every time he reminds his readers of it, which he does frequently in Deep Thinking. It is partly a memoir, partly a history of chess, computers and computer chess. It ends with an in-depth account of his return match in 1997 against IBM’s chess computer Deep Blue – Kasparov had beaten the computer the previous year – and a convincing argument that the match was not actually as historic as is claimed.
Chess mastery was once seen as a uniquely human form of superintelligence. Early artificial intelligence
(AI) researchers speculated that an algorithm capable of beating humans at chess could perform all the other cognitive tasks that make our species unique.
IBM’s marketing department endorsed this view, promoting Kasparov vs Deep Blue as an epic contest between humanity and the machines. And the machines won! The company’s share price soared! Kasparov threw a tantrum – “I am not a good loser,” he admits – and accused IBM of bad faith. He demanded a rematch, which the company refused.
Deep Thinking is good when describing the psychological drama of confronting an opponent that cannot know fear or make mistakes. It’s also good at deconstructing the mythology of the contest. Deep Blue was not a sophisticated form of machine intelligence, Kasparov argues. It was a triumph of hardware engineering and a “bruteforce” approach to computation. And it could not think creatively: its software was trained and optimised by chess grand masters secretly hired by IBM.
He’s optimistic, assuring readers that worrying about the loss of jobs is like complaining that antibiotics put gravediggers out of work.
The computer was dismantled after the match. It was good at beating humans at chess and generating publicity for IBM but not much else. Subsequent breakthroughs in AI came from companies such as Google and Facebook.
Like many technology enthusiasts, Kasparov believes that true machine intelligence – from self-driving vehicles to clinical diagnostic algorithms that outperform human physicians – is about to transform our world, rendering most careers obsolete. He’s optimistic about this change, assuring readers that worrying about the loss of jobs is like complaining that antibiotics put gravediggers out of work.
So the future looks bright, at least for the machines. Chess grand master Jonathan Rowson has even predicted that one day an artificial intelligence will completely solve chess, developing a strategy that will win or draw every game, “unless runaway global warming or nuclear war gets in the way”.
Maybe computer chess was not the most important problem to solve?
DEEP THINKING: Where Machine Intelligence Ends and Human Creativity Begins, by Garry Kasparov (John Murray, $37.99)