Hey there, time traveller!
This article was published 5/1/2013 (1302 days ago), so information in it may no longer be current.
RAY Kurzweil, the American inventor, artificial intelligence pioneer and futurist, is an optimist. For him, the brain half full.
Don’t worry about climate change, political upheavals, economic chaos. The future is bright. Because computers are learning to analyze information the way the brain does, artificial intelligence will change the world.
Our brains will mesh with computers. We will store our memories in the cloud, put "nanobots" in our brains to help us think, and thus maximize our potential.
Although How to Create a Mind is not an easy read, it is a valuable one. The chapters discussing how the human mind operates, what consciousness is, and how to differentiate between artificial intelligence and the human mind, are brilliantly illuminating.
Kurzweil uses Watson, the IBM super-computer that beat the two greatest Jeopardy! contestants, to show that current programming of computers is already different from the earlier computers that answered questions or played chess or checkers.
Whereas the earlier ones were given thousands of "if, then" instructions, Watson was not given the answers. Watson was provided with sets of learning algorithms and a great deal of hard data; it was then able to learn from those data so that it could understand questions in the forms of statements containing puns, and come up with the most probable answers in the forms of questions.
This clearly constitutes a quantum leap. Kurzweil believes that Watson is the harbinger of things to come.
Kurzweil’s thesis is not that the brain is complex, but that it is simple. There is, he writes, one "basic ingenious mechanism for recognizing, remembering, and predicting a pattern, repeated in the neocortex hundreds of millions of times."
This mechanism, which he describes as the "pattern recognition theory of mind," is, he says, exactly how the computer operates.
Thus melding the brain and the computer is simply a technological problem. It will be solved by 2029, given his predictive theory of technological advances.
The book is written with a light touch. "If you were at a cocktail party and there were both ‘normal’ humans and zombies, how would you tell the difference? Perhaps this sounds like a cocktail party you have attended."
Kurzweil takes us on a whirlwind tour of human thought, the anatomy of the brain, computer anatomy and thinking, human consciousness, and the future, with refreshing stops of jokes, stories and personal recollections.
Although the book is fascinating and well-written in a conversational style, the middle chapters make for a slow and difficult read because of the detailed discussion of neurology and computer language, and some of the later chapters are so dense with provocative insights that they require careful study. The majority of the many illustrations and charts in the book are also difficult to interpret.
Throughout much of the book, the layperson’s mind reels, overwhelmed by unfamiliar words, unfamiliar concepts, and discursive digressions into fascinating but obscure experiments that Kurzweil believes lay the groundwork for the work that has to be done to meld the brain and the computer.
Kurzweil might provide more information than many readers interested in his subject matter will want, and he might be riding a few hobby horses and want to answer every single objection brought against his theories, but what he has to say is extremely important.
If you give Kurzweil his assumptions, and skip the details, his thesis is fascinating and well-presented. So skim through the chapters that contain more information than you want, and study the chapters that contrast human consciousness and identity with how computers are learning to learn. You will be rewarded by a book filled with insights and provocative suggestions.
Lawrie Cherniack is a Winnipeg mediator and adjudicator who grew up on science fiction that has become science fact.