Monday, February 19, 2007

Intuition, Redefined?

"A clever person solves a problem. A wise one avoids it." -- Albert Einstein

When it comes to intuition, Einstein was always on-the-mark. What a great batting average.

Perhaps it's time to discuss the names of this blog. For instance, how do we boldly claim to "redefine" intuition?

Capyblanca, as in "", is a Hofstadterian invention, for an ambitious project in cognitive science, blending the ideas from The Copycat Project with Chess Champion José Raúl Capablanca.

Capablanca made perhaps the best remark, ever, about human intuition. While mentioning that other players tried to look ahead and see many moves in the evolving game, Capablanca assured us: "I see only one move. Always the best one."

Capablanca didn't bother to think through many alternatives, his gigantic, unique, experience in Chess enabled him to just "see the best move." If you think that was just plain arrogance and did not correspond to psychological fact, consider this: Ron "Suki" King, a world checkers champion, played in 1998 a simultaneous game against 385 opponents. He beat them all. Now let's make some calculations. Supposed he thought for two seconds in each position. A mere two seconds; time to look at it, have an initial idea, then make a movement. That gave each of his opponents 12 minutes and 30 seconds to consider the reply. So if you think that Capablanca was not telling the absolute truth about his own thought process, I'd love to hear an alternative explanation for the powers of Mr King.

Intuition is the ability to, as Einstein said about a wise person, avoid problems, even without thinking about them. Most decision-making theories, such as in Management Science, Cognitive Science, Game Theory, and so forth, have considered that decision is choice; and moreover, that choice selection is possible after evaluating the alternatives. But Capablanca and King were not evaluating alternatives; and this is our anchoring point in looking at the decision sciences.

Intuition is often defined as "immediate knowledge". Fair enough. But that definition needs, perhaps, some qualifications. I'd define it as "almost immediate situation understanding". The process is not immediate, of course, it takes a (little) time and (invisible, subconscious,) effort. And it's not about 'knowledge' as in a static form; it's about having a deeply meaningful state of mind, of understanding a situation very rapidly.

So here is our "redefinition". We have written about both the 'almost immediate' and the 'situation understanding' parts. About 'almost immediate', in "Minds & Machines", in "Artificial Intelligence", and in EuroCogSci2007. We have written about the "situation understanding" part in "Artificial Intelligence", in (to-appear-one-day) "Cognitive Science", and a host other papers still under review.

So, though I think these qualifications have sweeping implications for many theories in many fields, you might think that that "redefinition" is pretty mediocre. To that, I'd reply with a quote, by my great friend, Harry Foundalis:

"Alex, given the proper perspective, every single piece of work is bound to be mediocre."