Briefly put, Hawkins thinks intelligence is nothing more than memory, and the ability to predict.
Hawkins, who longs for the day when he and his colleagues will be able to produce an artificial "brain" that can think better than humans and not be encumbered with all the baggage, like emotions, that can distort our reasoning. He doesn't think these futuristic computers will ever look like us, or feel like us, or be the chummy robots shown in sci-fi flicks.
But they might be able to run our air traffic control systems, or explore other planets and do so with extraordinary intelligence.
But first, he says, computer scientists have to throw out much of what they have believed and adopt his theory. The human brain doesn't work like a computer. So scientists are on the wrong track if they think they're going to build a super fast computer that can think like Stephen Hawkings. Or even Barry Bonds.
It all came to him in April 1986 when he was sitting in his Northern California office, just thinking, of course.
"I was contemplating what it means to 'understand' something," he writes.
As he glanced around his office he saw things he was familiar with, furniture, books, that kind of stuff. And suddenly, he says, he had an "aha moment."
He "understood" his office. That's because his eyes fell upon a blue coffee mug that he knew, from personal memory, wasn't supposed to be there. And therein lies the essence of Hawkins' theory.
The human brain is filled with memories from previous experiences. That memory bank provides the basis for what we expect to encounter in the world around us. So if you look around the room in which you are sitting, you will see objects that your brain has already predicted will be there, based on its memories.
Intelligence really gets down to work when the prediction turns out to be wrong -- as in the case of a blue coffee mug that we suddenly realize shouldn't be there. Our eyes send messages to the brain, telling it that something is amiss, and the brain amends its memory bank and sends a message back down the chain of command, telling the observer what to do. Or perhaps more accurately, predicting what will happen if the observer takes a particular action.
