From newshub.ccs.yorku.ca!ists!helios.physics.utoronto.ca!news-server.csri.toronto.edu!rpi!batcomputer!cornell!rochester!yamauchi Sun Dec  1 13:05:39 EST 1991
Article 1651 of comp.ai.philosophy:
Xref: newshub.ccs.yorku.ca sci.philosophy.tech:1162 comp.ai.philosophy:1651
Newsgroups: sci.philosophy.tech,comp.ai.philosophy
Path: newshub.ccs.yorku.ca!ists!helios.physics.utoronto.ca!news-server.csri.toronto.edu!rpi!batcomputer!cornell!rochester!yamauchi
>From: yamauchi@cs.rochester.edu (Brian Yamauchi)
Subject: Re: Consciousness (was Re: Daniel Dennett)
In-Reply-To: zeleny@zariski.harvard.edu's message of 26 Nov 91 18:59:51 GMT
Message-ID: <YAMAUCHI.91Nov26184949@indigo.cs.rochester.edu>
Sender: yamauchi@cs.rochester.edu (Brian Yamauchi)
Nntp-Posting-Host: indigo.cs.rochester.edu
Organization: University of Rochester
References: <JMC.91Nov17135110@SAIL.Stanford.EDU> <1991Nov17.190935.5546@husc3.harvard.edu>
	<YAMAUCHI.91Nov26024948@indigo.cs.rochester.edu>
	<1991Nov26.135953.5926@husc3.harvard.edu>
Date: 26 Nov 91 18:49:49

In article <1991Nov26.135953.5926@husc3.harvard.edu> zeleny@zariski.harvard.edu (Mikhail Zeleny) writes:
>In article <YAMAUCHI.91Nov26024948@indigo.cs.rochester.edu> 
>yamauchi@cs.rochester.edu (Brian Yamauchi) writes:

>>In article <1991Nov25.101026.5866@husc3.harvard.edu> 
>>zeleny@zariski.harvard.edu (Mikhail Zeleny) writes:

>MZ:
>>>>>I'd like to support this view by
>>>>>noting that if you accept Colin McGinn's arguments that...

>BY:
>>What are McGinn's arguments?  Why on Earth should humans be inherently
>>incapable of understanding consciousness?

>I was going to summarize McGinn's argument, but I changed my mind.  If this
>is to be an informed conversation, it would be irresponsible of me to
>initiate another discussion of a philosopher no one among my interlocutors
>has read.

In any discussion between scientists and philosophers, it's generally
safe to assume that the scientists have read more of the scientific
literature, and the philosophers have read more of the philosophical
literature.  It seems a bit pointless to support your opinions based
on arguments that you are unwilling to explain.  One might be
interested in reading McGinn, if you could provide some indication
that his support for a seemingly unprovable assertion is based on
something other than rhetorical handwaving...

>Instead, I'll suggest a reductio ad absurdum of the AI view.  Assume that
>the mind is reducible to the functioning of the brain.  Then we may
>conclude that the mind shares the computational limitations of an FSA.
>Consider the fact that, in contrast with Turing Machines, there is no such
>thing as an Universal FSA.  

I always wince when formal automata are dragged into discussions of AI
philosophy.  Sure, the human mind can be viewed as an FSA, but what
does this prove?  An FSA can be built to recognize *any* finite
language, and humans are not commonly in the practice of recognizing
infinite languages (in the formal sense) because things like a finite
lifespan tend to get in the way.  Furthermore, if you allow machines,
like humans, to make educated guesses and occasional mistakes, the
usefulness of this analogy disappears completely.

It strikes me as being remarkably unenlightening to view human
intelligence in terms of formal automata and languages.  Automata are
abstract models of computation that are useful to theorists studying
various mathematical properties of formal computability and
complexity.  While complexity can be relevant to the feasibility of AI
algorithms, formal computability is almost completely irrelevant to
types of intelligence required for humans and animals to survive and
prosper in the real world.  These tasks are trivially computable in
the theoretical sense (enumerate all possibilities, build a lookup
table), but are extremely hard in terms of actual algorithms that will
operate on real-world hardware.
--
_______________________________________________________________________________

Brian Yamauchi				NASA/Caltech Jet Propulsion Laboratory
yamauchi@cs.rochester.edu		Robotic Intelligence Group
_______________________________________________________________________________



