Here's the full video: you can skip past the intro if you want to Edelman's lecture @ 19:00
Mad props to ScorpDiesel for the assist with the embed
[FLASH]http://video.google.com/googleplayer.swf?docid=7437432153763631391&hl=en&fs=true[/FLASH]
Link: http://video.google.com/videoplay?docid=7437432153763631391#
PowerPoint Presentation: http://www.almaden.ibm.com/institute/resources/2006/Almaden Institute Gerald Edelman.ppt
My comments:
1) His theory (TNGS) rejects dualism, reductive naturalism and the notion of the brain as an algorithmic computing machine ("Turing Machine"). Rather, he suggests that the brain works in a "context dependent" process via perceptual categorization which is adaptive (through neuronal group-selection). And through "re-entrant" feed-back mechanisms (which effectively modulate synaptic firing/connectivity) it generates responses to stimuli with respect to value systems/rewards.
The process of memory is dynamic and creative and NOT information-retrieving and replicative like how computers work. This shits on all digital physics matrix type brain-in-a-vat ideas of consciousness. Consciousness is modulated by attention and NOT by a programed (intrinsic or extrinsic) set of instructions.
2) His referencing of the concept of degeneracy in all of biology is brilliant! Check what he says from 1:06
"... ambiguity is powerful because interaction degenerate systems cause associativity, and we need that to start with ... metaphors begin ..."
3) What's impressive, to me, is that he's actually a developed an empirically founded scientific theory--not a hypothesis--that can be further developed as measurement efficacy improves.
4) When you think about it, his theory of neural Darwinism explains and helps understand why neurons are the only cells in the body that never renew.
5) Check out his exchange with the electrical engineer dude that asks the question @ 1:16:25. Edelman asks him if he believe evolution is a Turing Machine. Dude says he can write genetic algorithms that can 'simulate' effects that underlie evolution. Edelman pretty much ethers that point by saying that a) then you should be able to write a genetic program that can predict the out come of evolution in a million years. Can't be done. and b) you can't generate that program
ab intio without human involvement.
6) When you think about it, the "demon" of reductionism is that it inevitably eventually breaks down to "the first cause" problem. In other words, reductionism works for science up to the point where matter, or the idea, is irreducible. At which point a creationist theory is developed (Big-Bang, etc). This, imo, is why the quest for a TOE solely based on the reductionist approach is futile.