From newshub.ccs.yorku.ca!ists!helios.physics.utoronto.ca!news-server.csri.toronto.edu!utgpu!csd.unb.ca!morgan.ucs.mun.ca!nstn.ns.ca!bonnie.concordia.ca!uunet!tarpit!cs.ucf.edu!news Tue Mar 24 09:57:26 EST 1992
Article 4608 of comp.ai.philosophy:
Path: newshub.ccs.yorku.ca!ists!helios.physics.utoronto.ca!news-server.csri.toronto.edu!utgpu!csd.unb.ca!morgan.ucs.mun.ca!nstn.ns.ca!bonnie.concordia.ca!uunet!tarpit!cs.ucf.edu!news
>From: clarke@acme.ucf.edu (Thomas Clarke)
Newsgroups: comp.ai.philosophy
Subject: Re: Infinite Minds?
Message-ID: <1992Mar19.142419.9413@cs.ucf.edu>
Date: 19 Mar 92 14:24:19 GMT
References: <1992Mar17.211601.18028@neptune.inf.ethz.ch>
Sender: news@cs.ucf.edu (News system)
Organization: University of Central Florida
Lines: 63

In article <1992Mar17.211601.18028@neptune.inf.ethz.ch> santas@inf.ethz.ch  
(Philip Santas) writes:
| 
| In article <1992Mar16.165050.4693@cs.ucf.edu> clarke@acme.ucf.edu (Thomas  
Clarke) writes:
| >In article <1992Mar11.214547.28524@neptune.inf.ethz.ch> santas@inf.ethz.ch
| >(Philip Santas) writes:
| >|
| >| There is evidence that supports the digital functioning, and I do
| >| not see any reason why we should rush to abandon this theory BEFORE we
| >| get evidence for the contrary.
| >|
| >I am from the school that  we don't really understand something until we can
| >build it (maybe not even then).  When speech recognition and machine vision
| >reach human (animal even) levels of competence using purely digital  
techniques,
| >then I will accede ...
| 
| But systems like cinema, TV, CDs and whatever, are based on the attribute
| of our neural system to behave digitally. 
| 
| Of course there are some who simply close their eyes and expose their 
| prejudices, and wonder if flying exists before birds in order to prove
| that airplanes do not actually fly... 
| 
| On the other hand, there are scientists and engineers who try to accomplish 
| something with open mind, free from dogmatic beliefs,
| like Negroponte  in MIT, who is working towards squezing the whole work of 
| Bach into a single CD!
| 
| >Until that day comes, it remains a _philosophical_ possiblity that the brain
| >does not function digitally. 
 
The ability of the brain to make sense of a sampled series of data, as in  
cinema images, is truly wonderful.  Experiments seem to imply that the brain  
even maintains a running model of what is viewed. Experimental subjects seem  
able to makes responses in experiments that are easiest to understand if they  
have access to the positions of moving objects at positions intermediate  
between those shown in the frames.   Sorry for vagueness, but reference is not  
at hand - experiment was something like  cinematically viewing a moving point  
of light and then making judgements about its position relative to a second  
point of light flashed on in the time interval between cinematic frames.   
Discrimination was just as accurate as for DC/analog imagery.  Results are not  
accounted for by the persistence of vision.

I could code a system using Kalman digitial filtering etc, that could reproduce  
this performance for a well defined point of light, but for the case of  
arbitrary images, I can't think of any approach based on von Neumann/Turing  
machines that would not run afoul of extreme computational complexity problems.   
It seems only some sort of massively parallel approach will work.  In the brain  
this leads to consideration of pulse rate coded (analog) neural networks with  
all the complications of possibly chaotic dynamics, and also to (admittedly  
speculative) quantum considerations such as I outlined in my thought  
experiment.

At any rate, it seems to me that there is ample room for the possibility of non  
von Neumann/Turing modes of computation in the brain.  This would of course  
imply that the brain is in some sense non-recursive, and opens the way for  
violations of the Church-Turing thesis.  If the operation of the brain is  
non-recursive, then it may be the case that humans have effective means for  
non-recursive calculation at their disposal.  On the other hand, it may be the  
case that even non-recursive processes manifest themselves symbolically only in  
recursive ways.


