Newsgroups: comp.lang.prolog,comp.ai.philosophy
Path: cantaloupe.srv.cs.cmu.edu!bb3.andrew.cmu.edu!newsfeed.pitt.edu!gatech!news.mathworks.com!uhog.mit.edu!news!ml.media.mit.edu!minsky
From: minsky@ml.media.mit.edu (Marvin Minsky)
Subject: Re: please help me out with this!!
Message-ID: <1996Mar14.163730.4374@media.mit.edu>
To: mcgiml@hg.uleth.ca
Sender: news@media.mit.edu (USENET News System)
Organization: MIT Media Lab
References: <4i79o3$9a4@holly.cc.uleth.ca>
Date: Thu, 14 Mar 1996 16:37:30 GMT
Lines: 37
Xref: glinda.oz.cs.cmu.edu comp.lang.prolog:14976 comp.ai.philosophy:38922

In article <4i79o3$9a4@holly.cc.uleth.ca> mcgiml@hg.uleth.ca writes:
>Greetings all!

>	I have a query about the philosophical implications of prolog.  Namely,
>what is it about prolog that makes it such a good language to program
>expert systems and AI in?   (I guess I'm asking what aspects of prolog make
>it stand out from other computer languages).  Is it because prolog is an
>implication (if-then) based language, or are there other aspects of the
>language that are eluding me?

It is good for IF-THEN's because it's statements are basically
IF-THENs. That makes it good for programming some kunds of
"rule-based" expert systems.  It's also good for logic programming
because its primitives are logical connectives.  Both features make it
convenient for programming solutions to problems you already
understand rather well -- provided that they are easily represented in
predicate calculus.  Generally, my impresion is that prolog has not
been adequate for some, but not good for most other AI research,
because it (a) made it hard to apply knowledge to the search process
itself and (b) made it hard to use representations other than
predicate calculus.

I'm sure many prolog users will disagree with this.  However, the
disgreement may be less about the language itself than about the
value, in AI, of using predicate calculus as the base represntation.
So it's important to recognize that.  To put it diplomatically, "It's
not the language, stupid., it's the representation."  Prolog makes
them both almost the same.

>	Another question that I would like to have answered  (if you would be
>so kind) is how a program written in prolog is capable of making decisions
>(and simulating human intelligence).  Is this ability resident in the
>programming language itself, or does this ability come from the programming
>done by human experts?

That issue becomes vacuous for programs smart enough to figure
out new ways to solve problems.
