Review: Deep Thinking: Where Machine Intelligence Ends and Human Creativity Begins

Deep Thinking: Where Machine Intelligence Ends and Human Creativity Begins by Garry Kasparov
My rating: 4 of 5 stars

Garry Kasparov has a way of his own – it is well after his retirement from professional chess, and yet he is so much sought after. This book traces the progress of chess engines, the evolution of their algorithms culminating in the famous Kasparov Vs Deep Blue battle.

Chess engines have gained rapidly in the recent past from databases for reviewing games, to aids for analyzing positions, to challenging humans and finally overpowering them. Kasparov discusses how the algorithms have changed – the limits of technology resulting in lower search depth initially for brute force algorithms, to the incorporation of intelligent search tree reduction and later machine learning. While Chess has been of keen interest to me since years, the information presented is very systematic and well organized with quite a bit of information making for very interesting reading which I was not aware of.

Kasparov hates losing..well, he really does. Till a time it seemed to be smooth going against both human challengers and the machines. He is the world champion and quite easily overpowers Deep Thought – the predecessor to Deep Blue. He wins the first championship against Deep Blue as well, employing clever anti-computer strategies. The Deep Blue which he takes on in 1997 seems to be a different beast though. While Kasparov wins the first game, the second one turns out to be a stunner with Deep Blue employing strategies never before seen in chess engines. Kasparov, in fact becomes deeply suspicious, even alluding to the possibility of human intervention. He admits that the impact to his psyche as a result of this game was so high, that he subsequently loses the tournament. He is not exactly a gracious loser and has issues with not being given logs on time as well. Of course, subsequent revelations do indicate that IBM went to great lengths to prove a point and defeat Kasparov. This includes injecting some delays in Deep Blue to score some psychological points against Kasparov and tuning the opening book on and off. Kasparov mentions that Lou Gerstner visited the match area and gave the Deep Blue team a pep talk and said that the match was between the best chess player in the world and Kasparov.

This book is very well written and Kasparov's intellect really shines through. A large portion of the matter in the book is public information though, if you can make the effort to read and collate it. I really liked Kasparov's writing style and the only thing which could have been avoided was a few excuses for his intemperate behaviour at times such as being a sore loser to Deep Blue, and slamming the door in the game against Anand.

Kasparov's take on Artifical Intelligence / Robots in general are extremely mature and practical. In contrast to the general panic futurists have started spreading on impending large job losses, Kasparov has sensible comments to make and sound advice as well. As he points out - going into a panic on machines overpowering us at this point in time is like worrying about overcrowding on Mars! We are very capable of making a good future for ourselves if we are sensible about it.

If you follow chess, you will love this book!


View all my reviews