Book Review: Deep Thinking by Garry Kasparov
Book Review: Deep Thinking: Where Artificial Intelligence Ends and Human Creativity Begins
26 August 2017
Garry Kasparov. Many consider him history’s greatest chess grandmaster. In his latest work, Deep Thinking: Where Artificial Intelligence Ends and Human Creativity Begins, Kasparov flexes a different set of intellectual muscles, turning his brainpower from the tactics of the game board to humanity’s coming conflict with artificially intelligent machines.
Brutally objective on the inevitable dominance of artificial intelligence over many aspects of human society (“There is no back, only forward. We don’t get to pick and choose when technological progress stops …”) and his deeply personal defeat to IBM’s multi-million dollar chess supercomputer “Deep Blue” in 1997, Kasparov is a uniquely credible expert on the man versus machine debate.
After decades battling the machines, his optimistic outlook may come as a shock:
“We can either see these changes as a robotic hand closing around or necks or one that can lift us up higher than we can reach on our own... Instead of worrying about what machines can do, we should worry more about what they still cannot do.” - Garry Kasparov, Deep Thinking: Where Artificial Intelligence Ends and Human Creativity Begins
Kasparov’s advice to AI skeptics: “worry more about what they [the machines] still cannot do.” This can be interpreted in two ways.
The Rational Optimist Perspective
On one hand, Kasparov’s call to worry about what machines still cannot do is a call for them to do even more.
Kasparov observes that outsourcing tasks traditionally completed by humans to machines generally results in higher qualities of life, typically in the form of increased time and energy to focus on more desirable pursuits. This echoes the “Rational Optimist Theory” coined by British scientist Matt Ridley. Ridley demonstrated in The Rational Optimist: How Prosperity Evolves that, if the past is any indicator of our future, we should not fear advanced technologies which have historically produced longer, healthier lives for humans throughout the span of human history.
Applying Rational Optimist Theory to artificial intelligence, we ask if popular technologies like Google Maps or Waze make driving cars more efficient by automating human decision making (i.e. determining the fastest route from A to B), then why not delegate further authority to the machines, like actual operation of the car? Once machines are driving the car, why not automate all cars on the highway to fully optimize the entire traffic system?
At the point of maximum automation, all the humans in the all the self-driving cars can sit back, relax and think about a more important problem...or enjoy the ride. Your choice.
In my field of law, machines are learning how to complete tasks traditionally delegated to junior lawyers, like document drafting and contract review. Some can do more advanced work, like drafting memorandums of law, case law analysis and pre-transactional due diligence. I view this as a positive development, as this gives precious time back to the lawyers to solve the more complex and genuinely fulfilling problems that brought us to law school in the first place, like a vexing policy question or litigation defense strategy.
The Human Learning Perspective
On the other hand, Kasparov’s call to focus on what machines cannot do is a call to think harder about what humans can. This brings us to the unexplored, but closer-to-home realm suggested in Kasparov’s title: where artificial intelligence ends and “human creativity” begins.
Kasparov’s loss to Deep Blue was as painful as it was shocking. However, after years of reflection, the grandmaster concludes his defeat at the keys of Deep Blue was the ultimate learning experience, as it brought his own humanity into sharper relief.
Fascinatingly, by competing against an almost alienlike supercomputer, Kasparvo developed a deeper understanding of his own human play, including his human strengths (creativity, strategy) and weaknesses (narrative fallacies, psychological and emotional pressures). His distillation of the human mind from that of the machine is worth exploring, briefly.
Type A vs. Type B Thinking
Machines excel at “Type A” or “brute force” thinking. This is the ability to simply out-calculate the opposition by examining the value of all possible moves from a given position and select the most logical. This turns out to be useful in chess where there are 300 billion ways to play the first 4 moves and the total number of legal positions is roughly comparable to the number of atoms in the solar system.
In contrast, humans excel at “Type B” or “creative” thinking. Type B thinking is far less understood and appears to rely on our superior visual system and pattern recognition ability. This allows humans to “see” more generalized, strategic developments, by focusing on a few good moves (as opposed to all possible moves in a set) and looking deeply at those particular moves in context. In Kasparov’s eyes, Type B thinking elevated chess from a game to an art – a mode of personal expression as opposed to cold, mathematical truths.
Type B ruled for a while, but Kasparov unequivocally admits that 1997 marked the coming of a new age where Type A machines would reign supreme for the foreseeable future.
He goes farther, predicting Type A processing will one day be so powerful that machines will completely “solve” entire games. Scientists at the University of Alberta have nearly solved checkers, and Google is working on a solution for the ancient Chinese game "Go." The point at which a game is completely “solved” would seemingly make it futile to play. The machine would calculate the path to victory no matter the initial move.
This prompted me to tweet the following question to Kasparov: assuming two computers that have “solved” the game of chess square off in a tournament, who wins? Infinite loop? No reply as of yet.
A Human Learning Era In Disguise?
In Deep Thinking, Garry Kasparov makes many important, optimistic contributions to the man versus machine debate. As a Rational Optimist, Kasparov sees the potential for machines to unlock newfound time and energy for humans to focus on more important, more pleasurable pursuits. He also understands that machines may ironically serve as mirrors to ourselves, joining a chorus of futurists predicting the machine learning age will actually be “human learning” age in disguise.
Crucially, Kasparov wants us to understand that chess is just the first domino. We will soon build machines that will defeat us in other ways. Tasks will be delegated and jobs lost. However, the end result may be a future of yet unimagined and higher potentials. Some final thoughts from the grandmaster:
“I have argued that our technology can make us more human by freeing us to be more creative, but there is more to being human than creativity. We have other qualities that machines cannot match. They have instructions while we have purpose. Machines cannot dream, not even in sleep mode. Humans can, and we will need our intelligent machines in order to turn our greatest dreams into reality. If we stop dreaming big dreams, if we stop looking for a greater purpose, then we may as well be machines ourselves.” - Garry Kasparov, Deep Thinking: Where Artificial Intelligence Ends and Human Creativity Begins
I highly recommend this book for anyone looking for an optimistic and inspiring take on where AI may take us in the years to come.
Long Island, New York