I’ve heard Vernor speak before. Somehow, I’m quite skeptical about his claim that computers will be able to exceed our cognitive capabilities relatively soon. AI has been over-sold for a very long time, so the field still suffers from credibility issues. Nevertheless, it seems to me that merely throwing Moore’s Law at something as architecturally complex as the human brain is problematical. Each neuron in our brain is a whole lot more than an integrate-and-fire computational element. Each neuron represents a constellation of hierarchical biochemical and biophysical processes constrained by an orderly, yet unique morphology. There are something like 100 billion of these neurons and each of them as perhaps 10,000 connections to others. How you go from a digital computer to something approaching the human brain is daunting.
I suppose the ultimate question is: will a PC pass the Turing Test anytime soon.
Jim
Discoverers and Discoveries 