Newsgroups: comp.ai
Path: cantaloupe.srv.cs.cmu.edu!rochester!cornellcs!newsstand.cit.cornell.edu!news.tc.cornell.edu!news.cac.psu.edu!news.math.psu.edu!chi-news.cic.net!newsxfer2.itd.umich.edu!newsxfer.itd.umich.edu!news.mathworks.com!uhog.mit.edu!news!ml.media.mit.edu!minsky
From: minsky@ml.media.mit.edu (Marvin Minsky)
Subject: Re: How is AI going?
Message-ID: <1996Mar3.022422.16665@media.mit.edu>
Sender: news@media.mit.edu (USENET News System)
Organization: MIT Media Lab
References: <4h7a2q$rqh@portal.gmu.edu>
Date: Sun, 3 Mar 1996 02:24:22 GMT
Lines: 37

In article <4h7a2q$rqh@portal.gmu.edu> kunderco@osf1.gmu.edu (Kirt Undercoffer) writes:
>To clarify my previous posting:  it appears that the central
>issues of AI are knowledge representation, inference, learning,
>and pattern recognition.  Once a specific problem is posed (machine
>translation for example) then we can get at very specific questions.
>My question is basically what are the great questions of AI from the
>standpoint of knowledge representation in general, inference in general,
>ect.  The only question of this sort that comes to mind is the frame
>problem.  Are there 9 other questions of the same nature that could be
>posed on the same sort of overarching level?

You might be amused to look at my old paper, "Steps Toward Artificial
Intelligence".  It begins with, 

"The problems of heuristic programming--of making computers solve
really difficult problems--are divided into five main areas: Search,
Pattern-Recognition, Learning, Planning, and Induction.

"A computer can do, in a sense, only what it is told to do. But even
when we do not know how to solve a certain problem, we may program a
machine (computer) to Search through some large space of solution
attempts. Unfortunately, this usually leads to an enormously
inefficient process. With Pattern-Recognition techniques, efficiency
can often be improved, by restricting the application of the machine's
methods to appropriate problems. Pattern-Recognition, together with
Learning, can be used to exploit generalizations based on accumulated
experience, further reducing search. By analyzing the situation, using
Planning methods, we may obtain a fundamental improvement by replacing
the given search with a much smaller, more appropriate exploration. To
manage broad classes of problems, machines will need to construct
models of their environments, using some scheme for Induction."

There are many hard questions in each of the five sections.  Cetainly,
they're not all the same questions I'd use in a modern version of that
paper; I'd focus it more around such questions as how to learn to use
several representations together, how we do common sense reasoning by
constructing useful analogies, etc.
