On Nov 19, 2008, at 11:17 AM, Stephen A. Lawrence wrote:
[snip interesting argument]
There's a line there somewhere between things that are conscious and
things that are not, but there's no way to determine with any
certainty
*where* to draw it, because the concept of "consciousness" is entirely
outside the ken of modern science.
I would claim that this is a rather important hole in our current
knowledge base.
Interesting you should chose to point this out. I used the term
"confabulation" in my prior posts in the not well known artificial
intelligence context. When deterministic computer programs obtain a
result from a given set of inputs it is fairly easy to determine
why. When self training neural networks produce an output
determining why is not possible because the complexity is
unfathomable and because the network changes in response to its
environment. In the early days of neural networks I recall the word
"confabulate" was chosen to describe what AI researchers did in
explaining in anthropomorphic terms, after the fact, "why" a neural
network took some particular action or produced some particular set
of output. Such explanations inherently depend on the context of the
assumptions, experiences, and linguistic limitations of the observer,
which are all entirely irrelevant to the actual performance of neural
networks of such limited size. I think behavioral neural networks,
i.e. brains, include as inputs random variables, so confabulation has
even less meaning in the context of describing why or how a brain
produces a given output. I think the word confabulate really tells
us much about how we observe nature. We engage in "cogent
confabulation", describing things in a manner most consistent with
what we already believe, and this is a demonstration of our sometimes
very limited ability to see things as they are.
I attended a lecture in the 1960's by a psychologist who was
developing his "assumpto-therapy". He developed his theraputic
technique to handle the many behavioral problems he saw which didn't
have clearly prescribed therapies and which typically resulted in
extended psychoanalysis many patients could not tolerate or afford.
He based his work on the premise that many ordinary behavioral
problems are caused by the patient having ingrained a false premise
at some early age. The objective of his therapy then was, through
dialog, to identify the false assumptions causing the problems as
quickly as possible and re-condition the patient. This technique is
also adaptable to self-therapy. This was apparently in many cases
very effective and took much less time than full conventional
psychoanalysis. I don't think his approach was accepted, but probably
wouldn't know if it were. I suppose it is a branch of transactional
analysis. I can see why some therapists would reject it in that it
substantially reduces fees.
Best regards,
Horace Heffner
http://www.mtaonline.net/~hheffner/
