Path: utzoo!news-server.csri.toronto.edu!cs.utexas.edu!sun-barr!olivea!uunet!elroy.jpl.nasa.gov!usc!zaphod.mps.ohio-state.edu!pacific.mps.ohio-state.edu!linac!att!pacbell.com!pacbell!pbhyf!rsprice
From: rsprice@PacBell.COM (Steve Price)
Newsgroups: comp.ai.philosophy
Subject: Re: computer life?
Keywords: Survival, instincts
Message-ID: <8661@pbhyf.PacBell.COM>
Date: 4 Mar 91 17:55:40 GMT
References: <1991Feb27.134800.18153@news.larc.nasa.gov> <1991Feb28.190553.20519@msuinfo.cl.msu.edu> <1991Feb28.235517.20218@news.larc.nasa.gov> <1991Mar1.143222.29977@msuinfo.cl.msu.edu> <1991Mar1.205136.10670@news.larc.nasa.gov>
Reply-To: rsprice@PacBell.COM (Steve Price)
Organization: Pacific * Bell, San Ramon, CA
Lines: 47

In article <1991Mar1.205136.10670@news.larc.nasa.gov> kludge@grissom.larc.nasa.gov ( Scott Dorsey) writes:
>In article <1991Mar1.143222.29977@msuinfo.cl.msu.edu> dailey@buster.cps.msu.edu (Chris Dailey) writes:
>
>>                                                    Have you ever
>>written a cellular automata program with differing rules?  Even though
>>you knew what the rules were before you ran the program, it was hard to
>>predict what the outcomes would be like.  Similar with chaos systems
>>like the Mandelbrot set:  each point is easy to find, but a point X and
>>Y units away is [almost?] impossible to predict.
>
>Yes.  But since the system is calculable, it's behaviour can be predicted.
I read James Gleich's book on Chaos last year, and drew the conclusion that
many physicists and mathematicians today disagree specifically with this
assertion that in any non-trivial system behaviour can be predicted (with
the degree of accurracy that seems to be desired here).  The classic
example is weather -- all the inputs and processes are subject to sampling,
measurement (and even a wee bit of control) but the outputs vary widely
from predictions; no hope of accurrate long-range weather forecasting is
considered even theoretically possible (if I understood Gleich's point).

>It's just more difficult to predict, and the pattern which occurs is not
>obvious at all to the naked eye.  (Sorry if I tend to have a rather 
>deterministic, nineenth century point of view on the world, but I think
>it's a rather good one).  If it's almost impossible to predict, that just
>means that the tools for predicting it haven't been devised yet.
>
I heard a radio interview with the man who was awarded the Nobel Prize 
(partially) for the discovery of Quarks.  (My ignorance is showing here
since I can't recall his name -- but the other mental presences on the 
Net will surely compensate for my lack).  He made the point that the 
inability to deterministically predict the behaviour of complex systems
is NOT a result of our inaccurate sampling or modeling tools.  It is a 
fundamental fact or aspect of reality.  

Of course he was referring specifically to reality at the quantuum level, 
but there are profound implications for us at the macro level; these need 
to be considered in any AI discussion.  Remember the weather? (By the way,
it is raining in California, thank God!)  We specifically need to be on guard 
against our "common sense" notions of reality, which may often just be a 
"rather deterministic, nineteenth century point of view".


-- 
Steve Price 		UNIX: pacbell!pbhyf!rsprice	PHONE: (415)823-1951

We must live within the ambiguity of partial freedom, partial power, and 
partial knowledge.