Book review: Garry Kasparov's Deep Thinking: Where Machine Intelligence Ends and Human Creativity Begins

This book, more than anything else, is a fascinating look at the way human decisions influence artificial intelligence.

(We do understand this, don't we? That everything an AI does – at least for now – was both designed and programmed by at least one human?)

Deep Thinking is also an explanation of exactly what happened when Kasparov played Deep Thought in 1989, and how those games affected the way both Kasparov and IBM prepared for Kasparov's chess match with Deep Blue in 1996 and the follow-up match in 1997.

Kasparov's error – which he freely admits – was his assumption that he was playing an artificial intelligence that had been designed to play chess, with everything we currently know about chess programmed into it.

He was actually playing an artificial intelligence that had been specifically designed to beat Garry Kasparov.

The idea that an individual AI can be programmed to manipulate an individual person's behavior has a lot of negative implications, many of which you can figure out on your own.

However, it also has a lot of positive implications, many of which have to do with our ability to learn more efficiently.

The 1:1 tutoring relationship, after all, still remains one of the most successful ways of helping a student understand how to solve problems – or, as we like to say in our house, "how to move from guessing to knowing."

Kasparov doesn't go so far as to say "let's recreate the Primer from Neal Stephenson's Diamond Age," although that is the logical endgame.

He does say that the next best move might involve not only AI but also IA – intelligence amplification.

Teach the machines to learn, the best way we know how.

Information, calculation, iteration, connection.

Then let them teach us.