Path: utzoo!attcan!uunet!seismo!ukma!wuarchive!decwrl!adobe!adobe.COM!asanders
From: asanders@adobe.COM
Newsgroups: comp.ai
Subject: Re: Is simulation what we need ?
Message-ID: <1877@adobe.UUCP>
Date: 22 Feb 90 19:04:19 GMT
Sender: news@adobe.COM
Lines: 34


>The question is:
>
>Is simulating thinking really thinking ?
>
>With regard to flying, simulating flying is not really flying. At least not
>to me. Where do we draw the line between real and simulation events ? So
>simulation does not convience me. All responses welcomed.

                                             -SURAJ C. SURENDARKUMAR.


In my opinion, it is necessary to define precisely what we hope to gain
from the "simulation" of phenomena in order for those simulations to be
meaningful. Simulated flying is NOT real flying, but -- provided we
understand flying sufficiently to model it accurately *for our purposes* --
we may succeed in creating a simulation that teaches potential fliers, or
perhaps designers of airplanes, something useful about flying.

I am not so certain that the analogy can be extended to thinking. While
flying is an essentially mechanical phenomenon based on well-understood
physical laws, *thinking* is an ephemeral process: as soon as you think you
have grabbed it, it pops up somewhere else -- like reflections in a hall of
mirrors. So, not only is simulated thinking NOT thinking; our understanding
of real thinking is so deficient that it is doubtful we can simulate it
accurately enough to learn anything useful about THINKING. This is not to
say that such research is without results. Obviously it is not. But those
results have their own reality and their own applications. The invention of 
artificial flowers gave rise to a new industry also; but no matter how
detailed, how carefully made an artificial flower might be, it will never
resemble a real flower *in its totality*. There is no reason why it should,
but claims to the contrary would surely be misleading.

Alan