Author: Miguel A. Ballicora
Date: 13:09:59 08/05/02
Go up one level in this thread
On August 05, 2002 at 15:11:44, Robert Hyatt wrote: >On August 05, 2002 at 12:05:56, Louis Fagliano wrote: [snip] >> Early on in the evolution of life it >>was a clear advantage if a bacterium reacted appropriately to an external >>stimulus and the way to do that was by pattern recogintion (holoistically) >>rather than digitally on/off. Pattern recognition is very hard to achieve if >>done digitally. Actually, the way that bacteria recognize and follow an stimulus is a combination of digital and analogic signals, and it depends how you want to look at it. Technically is neither, it is more of a probabilistic response of receptors. >I don't believe that at all. IE You define the pattern you want to recognize, >I will write a program to do it. And given enough processors to do the >recognition in parallel (as the human mind does it) I can probably do it >faster. And more accurately. The more accurate you are, the farther you are from natural intelligence. Intelligence is the ability to take data and make generalizations about it. The way to do it is through inaccuracy. An animal learns that a lion is bad for its health, and immediately knows that most probably a leopard will be bad too. A leopard is not the same as a lion but they look very similar. The more accurate the animal is in recognizing lions, the more will be eaten by leopards :-) At a human level, that inaccuracy is what allow us to make come up with ideas by association etc. Miguel
This page took 0 seconds to execute
Last modified: Thu, 15 Apr 21 08:11:13 -0700
Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.