Path: utzoo!utgpu!utstat!jarvis.csri.toronto.edu!mailrus!tut.cis.ohio-state.edu!ucbvax!agate!helios.ee.lbl.gov!nosc!humu!uhccux!lee
From: lee@uhccux.uhcc.hawaii.edu (Greg Lee)
Newsgroups: comp.ai
Subject: Re: Question on Chinese Room Argument
Message-ID: <3395@uhccux.uhcc.hawaii.edu>
Date: 6 Mar 89 19:50:12 GMT
References: <7431@polya.Stanford.EDU>
Organization: University of Hawaii
Lines: 16

From article <7431@polya.Stanford.EDU>, by geddis@polya.Stanford.EDU (Donald F. Geddis):
" ...
" I agree that complexity arguments are not important for this thought
" experiement. ...

I doubt you should agree.  What kind of "non-symbolic" computations
can't be modeled by symbolic ones?  Those which would require
too many symbols or rules, I suppose.  If that is the only
difficulty, complexity is an important issue.  Then to argue
the (in)feasibility of a life-like robot we would need to estimate
the number of "distinct" brain-states that could be caused
by perception, to establish at least whether it is finite.  This
seems to me to be the most plausible direction to pursue to
salvage Harnad's arguments.

		Greg, lee@uhccux.uhcc.hawaii.edu