Submitted by Paracuzia (not verified) on Tue, 03/16/2010 - 02:49.
I think we already have AI which is particularly good at specific things (Youtube's auto-caption is pretty cool; Wolfram Alpha's another example).
The reason I think we should find out what makes us intelligent, is that the neocortex seems to use the same "algorithm" for all types of data processing. That is, the neocortex processes visual data, auditory data, etc. in exactly the same way (or that's what it looks like). Maybe if we figure out what's going on, we'll be able to apply this general intelligent algorithm to an array of different problems.
Right now, context-specific AI is just that: context-specific. We're using very specialized algorithms for each different intelligent application. We still stand to learn A LOT about the brain, and I wouldn't be suprised if what we do find will be directly applicable to AI. That being said, I don't think AGI depends on the reverse engineering of the brain; there are other ways of achieving our goal.
I think we already have AI which is particularly good at specific things (Youtube's auto-caption is pretty cool; Wolfram Alpha's another example).
The reason I think we should find out what makes us intelligent, is that the neocortex seems to use the same "algorithm" for all types of data processing. That is, the neocortex processes visual data, auditory data, etc. in exactly the same way (or that's what it looks like). Maybe if we figure out what's going on, we'll be able to apply this general intelligent algorithm to an array of different problems.
Right now, context-specific AI is just that: context-specific. We're using very specialized algorithms for each different intelligent application. We still stand to learn A LOT about the brain, and I wouldn't be suprised if what we do find will be directly applicable to AI. That being said, I don't think AGI depends on the reverse engineering of the brain; there are other ways of achieving our goal.