Newsgroups: comp.ai.philosophy
Path: cantaloupe.srv.cs.cmu.edu!bb3.andrew.cmu.edu!news.sei.cmu.edu!cis.ohio-state.edu!math.ohio-state.edu!howland.reston.ans.net!ix.netcom.com!netcom.com!kovsky
From: kovsky@netcom.com (Bob Kovsky)
Subject: Re: Robot autonomy (was Is the mind/brain deterministic?)
Message-ID: <kovskyCzIvqt.CHn@netcom.com>
Organization: NETCOM On-line Communication Services (408 261-4700 guest)
References: <kovskyCzF8D4.Bxv@netcom.com> <3aghdc$q3s@cantaloupe.srv.cs.cmu.edu> <kovskyCzGyMx.ELv@netcom.com> <jqbCzHKJC.B6L@netcom.com>
Date: Sat, 19 Nov 1994 16:13:41 GMT
Lines: 97

In his posting, Mr. Balter juxtaposes several statements I have made in 
this thread to demonstrate aprarent contradictions.   For completeness, I 
copy his entire posting despite its length.

In article <jqbCzHKJC.B6L@netcom.com>, Jim Balter <jqb@netcom.com> wrote:
>In article <kovskyCzGyMx.ELv@netcom.com>, Bob Kovsky <kovsky@netcom.com> wrote:
>>	My skepticism about the possiblity of self-navigating robotic
>>travel was based on my understanding of current technology, not on
>>principle.
>
>Oh really?  Somehow your statements 
>
>> In earlier posts, Prof. Moravec and I have been exchanging badinage in a
>> debate where I contended that one reason AI has not achieved its promised
>> breakthroughs has been its turning away from study of biological function
>> and its insistence on the universal applicability of computational models.
>      ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
>
>> That you and your associates need 
>> to invest such enormous resources in order to accomplish something that 
>> does not even perform as well [as a fly] suggests that there is 
>> something missing in
>> ^^^^^^^^^^^^^^^^^^^^ 
>> your approach.
>  ^^^^^^^^^^^^^
>
>>	My sarcastic remark about "world-shaking breakthroughs" stands.  
>> After 25 years and billions of dollars, AI has failed to produce.  Other 
>> than some useful but minor "expert systems" operating in extremely 
>> limited domains, there is very little evidence that the model of 
>                   ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
>> universal mechanical computation has any validity.  And very little 
>  ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
>> return on the investment.
>
>seem to speak to some principle or another.
>
>-- 
><J Q B>

	When I used the phrases "universal mechanical computation" and the 
"universal applicability of computational models," I had in mind the 
notions, often expressed by AI people and others of a similar inclination 
that "cognition is computation," that  consciousness can be "explained" 
(a la Dennett) in terms of machines or "virtual  machines," and that 
computers can on principle, and will in the foreseeable future, closely 
mimic human intellectual functions.

	In my view, consciousness and many intellectual functions involve 
co-ordination of activities where the underlying structural systems are 
"incongruent," i.e. the structural systems do not fit together in any 
clear or simple way.  For example, when we describe in words a visual 
image in order to accomplish a particular purpose, we must co-ordinate a 
one-dimensional string of words governed by syntactical rules with a 
three-dimensional visual image governed by the geometry of space and of 
an inexhaustible richness in a context constrained by the purpose.  This 
task is, in my view, on principle beyond the reach of computation.
                     ^^ ^^^^^^^^^    
	Self-navigating robots involve the co-ordination of two functions,
vision and action, but the underlying structural systems are congruent in
that both functions are grounded in ordinary space.  Certainly, there is
nothing in principle that makes navigation on hard, stable surfaces beyond
the reach of machines.  I also see nothing inherently impossible about
navigation on unstable surfaces when there is sufficient information to
enable a proper course of action to be determined mechanically.  I would
have to think deeper about problems involving texture to be ascertained
only from visual "cuess", unforeseen conditions and imperfect information. 

	However, even in the case of visually-controlled navigation, my
view is that machines and humans are operating under different principles. 
Moreover, that there are clearly many intellectual functions performed by
humans that cannot be mimicked mechanically (or even performed
competently). 

	When I wrote about the "world shaking breakthroughs" that had 
(not) come from Prof. Moravec's laboratory, I was not aware of the the 
exact nature of his work and understood from his other postings in this 
conference that he was advocating a theory of "universal applicability of 
computational models."  Certainly his work is substantial and to the 
extent I was poorly informed, I must and do apologize. 

	Nonetheless, I stand by my earlier remark that AI (in the sense 
that it asserts "cognition is computation" etc.) has failed to produce 
results that evidence the truth of the assertion.  Thus my reference 
to the few "expert systems" that have proved pactical only in limited 
domains.

	I also stand by my reference to the capacity of a fly to navigate 
just about as well as robots guided by 100 MIPS computers.  And I stand 
by my opinion that those who ignore the capacity of biological neuronal 
systems to function quickly and competently on the basis of small and 
slow cognitive units are missing something important.
-- 

*   *    *    *    *    *    *    *    *    *    *    *    *    *    *    *   * 
    Bob Kovsky          |  A Natural Science of Freedom 
    kovsky@netcom.com   |  Materials available by anonymous ftp
                        |  At ftp.netcom.com/pub/freeedom
*   *    *    *    *    *    *    *    *    *    *    *    *    *    *    *   * 
