From newshub.ccs.yorku.ca!ists!helios.physics.utoronto.ca!news-server.csri.toronto.edu!rpi!usenet.coe.montana.edu!ogicse!psgrain!percy!nosun!hilbert!max Tue Nov 26 12:32:49 EST 1991
Article 1620 of comp.ai.philosophy:
Xref: newshub.ccs.yorku.ca sci.philosophy.tech:1139 comp.ai.philosophy:1620
Path: newshub.ccs.yorku.ca!ists!helios.physics.utoronto.ca!news-server.csri.toronto.edu!rpi!usenet.coe.montana.edu!ogicse!psgrain!percy!nosun!hilbert!max
>From: max@hilbert.cyprs.rain.com (Max Webb)
Newsgroups: sci.philosophy.tech,comp.ai.philosophy
Subject: Zeleny's argument DOA (Was Re: Daniel Dennett)
Summary: ANN behavior often impossible to analyse. You lose.
Message-ID: <1991Nov26.004802.1394@hilbert.cyprs.rain.com>
Date: 26 Nov 91 00:48:02 GMT
Article-I.D.: hilbert.1991Nov26.004802.1394
References: <1991Nov17.190935.5546@husc3.harvard.edu> <DAVIS.91Nov24033509@passy.ilog.fr> <1991Nov24.124945.5834@husc3.harvard.edu>
Organization: Cypress Semiconductor Northwest, Beaverton Oregon
Lines: 36

In article <1991Nov24.124945.5834@husc3.harvard.edu> zeleny@zariski.harvard.edu (Mikhail Zeleny) writes:
>[Assume we can build a machine that exhibits our behavior and competence].
>However, at any given time, by assessing its construction, we may comprehend
>all causal factors that influence its behavior (to the extent that this is a
>machine constructed by ourselves, I assume that we can do so, retracing, if
>necessary, the modifications imposed on the initial configuration by the
>learning process).

You assume wrongly. Even now, the behavior of only moderately complex
ANNs is often impossible to understand by looking only at the modified
weights (or even the history of those weight changes). I strongly suspect
that the more sophisticated and complex ANNs to come will be even harder
to analyse by looking at the encodings. Your argument dies right here.

You, as a philosopher, are trying to determine what is possible and
impossible from first principles - perhaps you can tell me why such analysis
of an ANN is so often completely intractable. If you can't, maybe you should
open yourself up to some real world experience - and work with a NN simulator.
Maybe, just maybe, you have accidently enshrined false assumptions in
your 'first principles', and then in your discourse of 1000 steps, veered
further and further from the truth as a result.

Having seen so many brilliant philosophers do just that, over and over
again, I hope you will pardon us if we just don't throw up our hands and
give up in the face of your formidable logic.

BTW, it is no less ignorant of you to try demarcate the limits of what ANN's can
do, without ever having tried one (or without ever having studied them,
apparently), than it is for someone else to discuss Searle's Chinese Room
argument having read only other peoples summaries.  (I will not call you the
names you apply to others. I will say that I consider your style obnoxious
and arrogant, though occasionally entertaining.)

You work from information too far removed from reality.

	Max


