Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!wuarchive!julius.cs.uiuc.edu!ux1.cso.uiuc.edu!ux1.cso.uiuc.edu!m.cs.uiuc.edu!gillies
From: gillies@m.cs.uiuc.edu
Newsgroups: comp.ai
Subject: Re: What Has Traditional AI Accomplishe
Message-ID: <3200031@m.cs.uiuc.edu>
Date: 17 Oct 90 07:00:00 GMT
References: <69367@lll-winken.LLNL.GOV>
Lines: 29
Nf-ID: #R:lll-winken.LLNL.GOV:69367:m.cs.uiuc.edu:3200031:000:1326
Nf-From: m.cs.uiuc.edu!gillies    Oct 17 02:00:00 1990


> We could add to the list a host of medical instruments such as
> cat-scans (sp?) and Ultra-sounds. (Both look at internal organs
> without surgery) Then there are tools such as character readers, text
> and graphic scanners, and the nifty little barcode readers at the
> supermarket.

It is wrong to classify individual problems as "AI".  I deeply resent
this when AI researchers commit this error, as the person above has.
The work on Catscans and NMR/MRI is good physics and a good application
of known methods of exact scientific computation, period.

Either AI is a scientific method of making computers solve problems or
it is nothing at all.

> Our definition of AI is basically "whatever we haven't figured out how
> to do yet."  As soon as AI research refines the methods, the problem
> falls out of the AI category.

Once again, the person who wrote the above quote has made an error in
judgement.  If a refinement of an AI method continues to solve a
problem (and no other methods produce superior results), then it is a
triumph of AI.

I believe symbolic intergration programs such as Mathematica are good
examples of refined AI computation.  Chess programs such as the
leading chess program from CMU (Deep Thought?) are bad examples since
they demonstrate that brute force triumphs over sophisticated
heuristics.