Hacker News Reader

http://lesswrong.com/lw/tx/optimization/

http://lesswrong.com/lw/va/measuring_optimization_power/

Eliezer Yudkowsky described this pretty well.

I went to the Wikipedia article

http://en.wikipedia.org/wiki/Jeff_Hawkins#Neuroscience

to make sure you were referring to the author whose book I have perused before. I will note for the record that Hawkins has not won wide acceptance for his view of intelligence among other researchers to date.

That sounds like a good proposal on a definition of intelligence, but it doesn't tell you why someone would be good at predicting something, which is what my point was: that when you try to claim that any one thing is the reason for someone's ability to, in this case, predict something, you're ignoring the multitudes of other reasons that are all interrelated.

All the various forms of intelligence can be seen as a form of predictive capability. For example, coordination of many elements or one's own body parts involves making moves which avoid chaos, collision, or some other dysfunction. Avoidance is a kind of prediction.