[Date Prev][Date Next][Thread Prev][Thread Next]   [Date Index] [Thread Index] [Author Index

The Goertzels and the Chaitins of A.I.Theory.....



I've done a little research into some of the anti-Rosennean AI enthusiasts like Ben Goertzel and Greg Chaitin, spurred on by Aloisius' deconstruction of Goertzel's argument. I've met Greg Chaitin, when at a conference in Santa Fe, with my father, in the early 1990's. My assessment of Chaitin is that he's been completely swallowed by the machine metaphor. He's all math and engineering with too little knowledge of biology and he's convinced himself that biology is expendable. OK. Whatever blows your hair back. I think he should go right on trying as hard as he can to do it his way. As my father used to say about people like that; it keeps them out of trouble.
 
Ben Goertzel is different. I find him vastly more interesting because his knowledge base is far wider and his curiosity is still way bigger than his arrogance-- as far as I can tell, anyway. He's got a fascinating imagination. But I think he's accepted a few too many preliminary conclusions too soon, which means he's stopped questioning them. If so, then he's doing what I diagnosed as the main problem of AI enthusiasts: He's forgetting the human factor. The reason Artificial Intelligence looks real to us is precisely because its a bunch of chunks of real intelligence-- taken from human thought processes and plugged into a machine capability. The question all AI enthusiasts seem to be asking is; When do you have enough transplanted pieces of human intelligence to make it a complete and independent artificial intelligence?
 
If my father's gut instinct is correct, the answer is: Never. You can't reach infinity by accretion. Science in the reductionist mode cannot create life it can only manipulate life. So it can splice DNA and transplant organs and breed weird hybrids and so on. Computing in the reductionist mode is doing the same exact thing, with intelligence or consciousness; splicing and transplanting. But reductionism can never approach or understand complexity because it destroys the main aspects of a complex system by reducing it; the organization and the causal impacts that the organization generates. It's the organization that brings infinity with it. You can't even study it if you've destroyed it before you begin! Conscious intelligence is a complex system within a complex system within........ and they're transplanting pieces of it into a simple system; a machine.
 
But hey, if you can make it mimic human intelligence convincingly enough, you'll probably get that grant money.
 
Judith
 
Tim Gwinn wrote:
Ah! That is a marvelous way to view it! Yes, yes - the human mind closes the
loop and breaks off the incipient infinite regress of syntactic referents in
the software; but only when the two are taken as a complete system: the
"chimerical Unum"!  Very nice!

Regards,
Tim
 
Judith Rosen wrote:
One thing AI enthusiasts keep forgetting is the human factor; what my father
referred to as the "chimerical system of human-and-computer or
human-and-internet". When people are using the internet and/or computational
technology, there really IS an "artificial intelligence" created by the
combination. No single human being would be capable of the kind of knowledge
base or speed of thought or global reach that these technologies give us.
(Although Google reminds me a lot of my father in it's ability to answer
questions and bring up all kinds of ancillary subjects of interest...) We
use technologies to augment our natural capacities, which turns us into
various chimerical systems of human/technology.

So, with the Semantic Web up and running, and human beings manipulating it
to extract information that they need for their own pursuits, THERE'S the
entailment that was lacking in the internet as a separate, discrete system.
The human mind closes the entailment loop. The Semantic Web is a Life
Sciences project-- in order that scientific data will be searchable and
retrievable in a much more precise and comprehensive way. The capability
that will give individual human minds is going to be huge.

The whole concept of E Pluribus Unum is one of pooling resources into one
really versatile, creative, capable chimerical system. The internet gives
each individual the opportunity to turn themselves into a chimerical
"Unum"...