LAWS@SRI-AI.ARPA (06/08/85)
From: AIList Moderator Kenneth Laws <AIList-REQUEST@SRI-AI>
AIList Digest Saturday, 8 Jun 1985 Volume 3 : Issue 76
Today's Topics:
Queries - Expert System-Based Databases &
Military and Financial Expert Systems & Technophobia,
AI Tools - Lisp on the Sun,
Humor - Capitalization,
New Journal - Applied Artificial Intelligence,
Obituary - King-sun Fu,
Philosophy - Self-Reproduction and Consciousness
----------------------------------------------------------------------
Date: Thu, 6 Jun 85 16:44 EDT
From: susan watkins <chaowatkins@SCRC-STONY-BROOK.ARPA>
Subject: expert system based database
I'm interested in finding out whether database systems using expert
system technology (or other AI techniques) have been developed.
Is it, for example, fruitful to think about the difference between an
expert-system-database and a conventional database with an `intelligent'
user-interface front end (presumably holding the `reasoning' part of the
knowledge that are domain specific to the application) ?
[The 1st Int. Conf. on Expert Database Systems will be held in
Charleston, April 1-4, 1986. See AIList vol. 3, no. 57, for an
announcement and list of the program committee. -- KIL]
------------------------------
Date: 7 Jun 1985 20:56-EDT
From: RSHU@BBNG.ARPA
Subject: Military and Financial Expert Systems
I noticed that the Intl. Workshops on Expert Systems and Applications will
include two areas which I am interested in. The first is J.F. Gilmore's talk
on military applications of expert systems. The second is Kershberg &
Dickinson's talk on an expert support system for financial analysis.
I would appreciate the following:
1) Pointers to the above mentioned researchers. Net addresses are preferred
physical mail would also be appreciated.
2) Pointers to anyone else working in either of these two fields.
I am currently working on a support system for tank platoon tactics. Our work
centers around reasoning with a terrain database to make conclusions about
visibility and trafficability. This project is not classified so I'll gladly
provide more information if anyone is interested.
------------------------------
Date: Fri 7 Jun 85 17:20:23-PDT
From: MARCEL@SRI-AI.ARPA
Subject: Love-Hate Computers
At present, computer science and especially artificial intelligence work is
very highly regarded by those outside the field. When people find out what
my trade is I can sometimes watch the points go up on their prestige score-
board. On the other hand there is something about computers and AI that can
arouse strong instinctive fears. For the most part it's a matter of speculation
what those fears are. Few people know the field well enough to judge it
accurately, and those who do know the field (and its resemblance to alchemy) no
longer fear it. That, and the fact that the articulated reason for fear is
unlikely to be the actual reason, makes it largely a matter of guesswork as
to what the problem is. Let me propose a few possibilities:
- AI threatens to make machines autonomous
- it's humiliating that mere machines can be smarter than us
- AI hopes to make machines capable of improving their own techniques,
thus automating the invasion of technique over spontaneity, and
multiplying Future Shock
- people don't know how to use computers, so feel intimidated
- people worry about being controlled by/dependent on machines
- we have to redefine what it is that makes us humane -- some kinds
of intelligence will become unimportant, and people are just
emotional machines -- so we lose some aspect of "what's special
about me"
- we may be machines ourselves, devoid of free will/responsibility
- machines will take over our jobs
I'd like to know of:
- other causes of technophobia
- references to literature analyzing technology (I know about Asimov,
Weizenbaum, Dreyfus, Dennett, Turkle, Ellul, Toffler, Wiener, Huxley
and the like)
Please send replies directly to me:
Marcel@SRI-AI.ARPA
------------------------------
Date: 7-Jun-85 10:08:49-PDT
From: mhb@FORD-WDL1.ARPA
Subject: Lisp on the Sun
Regarding query concerning Lisp on Sun, I am familiar with the
following three versions.
(1) Franz Lisp as part of the Berkeley Distribution Tape
(2) Franz Lisp sold by "Franz, Inc." in Berkeley
This has some Common Lisp features added to it.
Vendor: (415) 540-1224
(3) Common Lisp, currently under development by "Lucid" in Palo Alto
Phone: (415) 424-8855
Mike Bender
Ford Aerospace
------------------------------
Date: Fri 7 Jun 85 09:30:24-CDT
From: David Throop <LRC.Throop@UTEXAS-20.ARPA>
Subject: Capitalization: Various Entities
Class Capitalization Pronoun Capitalization
=============================================================================
Inanimate Object Normally lowercase Normally lowercase
(bugs, desks, toaster) (it, them)
People First letter Normally lowercase
(Tom, Dick, Harry) uppercase (him, her, they)
Dieties First letter Always uppercase
(Christ, Krishna, Athena) uppercase (He, Him, Her)
Expert Systems All letters (Answer uncertain,
(EMYCIN, ISIS, SOLOMON) uppercase ask One)
=============================================================================
[Aside from the humor in this progression, there seem to be at least
the following factors at work: 1) FORTRAN, COBOL, and other relics
of the golden age of punched cards; 2) copyright/trademark conventions;
3) acronym syntax; and 4) the right of a developer to name a system
whatever he wants. I have been told by an editor that any nonacronym
of more than five letters (4? 6?) should have only a single capital,
e.g., Prospector, but I think that rule 4 should take precedence.
Personally, I like all caps (except for AIList). -- KIL]
------------------------------
Date: Fri, 7 Jun 85 14:42:48 edt
From: rada@nlm-mcs (Roy Rada CSB)
Subject: New Journal - Applied Artificial Intelligence
Perhaps you too have recognized the same need that we, a
group of American and European AI researchers and practi-
cioners have felt regarding the absence of a journal devoted
to practical information such as
- applications of AI, e.g. use of expert systems, natural
language systems, speech, vision, and robotics, for
solving tasks in industry, management, administration,
and education,
- evaluations of existing AI systems and tools, espe-
cially comparative studies,
- user experience,
- theoretical research with relevance to potential appli-
cations,
- economic, social, and cultural impacts of AI.
An established publisher has joined our forces, and we are
now happy to announce:
AAI Applied Artificial Intelligence,
an International Journal.
The first issue of AAI will appear in January, 1986.
If you are interested in
- being informed about the subscription prices, which are
drastically reduced for members of AAAI, ACM, IEEE, and
of member organisations of ECCAI and IFSR,
- submitting a paper and want the Information for Authors
(accepted papers will be published rapidly--at least
now!),
please contact
Hemisphere Publishing Corporation
79 Madison Avenue, New York, NY 10016.
June 7, 1985
------------------------------
Date: Fri 7 Jun 85 09:43:18-PDT
From: Ken Laws <Laws@SRI-AI.ARPA>
Subject: Obituary - King-sun Fu, 1930-1985
[Reprinted from IEEE Computer, June 1985.]
King-sun Fu, Goss Distinguished Professor of Engineering and Professor
of Electrical Engineerinr at Purdue University died April 29 in
Washington, DC.
Fu was born October 2, 1930, in Nanking, China. He earned a BS degree
in electrical engineering in 1953 from National Taiwan University, an
MA in science in 1955 from the University of Toronto, and a PhD degree
in 1959 from the University of Illinois.
Fu was a member of the National Academy of Engineering, a member of
Academia Sinica, and a Guggenheim Fellow. He was the first president of
the International Association for Pattern Recognition. He received the
IEEE Award of Service in 1971, ASEE Senior Research Award in 1981,
IEEE Educational Medal in 1982, and AFIPS Harry Goode Memorial Award in
1982. He authored four widely adopted textbooks and edited more than
15 books, was author or coauthor of over 100 journal papers and over
200 conference papers. He also produced over 60 doctoral students.
He was a distinguished member and dedicated officer of the IEEE Computer
Society. He was elected a fellow of the IEEE in 1971 for his contributions
to the field of pattern recognition and leadership in engineering
education. He was the first editor-in-chief of the IEEE Transactions
on Pattern Analysis and Machine Intelligence and served as vice president
for publications for the Computer Society in 1982-83. He received several
awards from the Computer Society, including the Outstanding Paper Award
in 1973, Honor Roll in 1973, Certificate of Appreciation in 1977 and 1979,
and Special Award in 1982.
------------------------------
Date: Fri, 7 Jun 85 11:51:44 edt
From: FRAWLEY <20568%udel-cc-vax1.delaware@UDEL-LOUIE.ARPA>
Subject: Self-Reproduction and Consciousness
I'd like to make a few comments on the discussion about self-reproduction
and consciousness/intelligence.
As several people have pointed out (and I agree), size does not
ensure self-reproduction. This, of course, depends on what is meant
by "self-reproduction," but, Fodor's arguments against paramecia not
having mental representations notwithstanding, it looks as if
there is some evolutionatry continuity in consciousness: see
Griffin's book, The Question of Animal Awareness. Thus, size is not
the issue. After all, computers are very large, but they have
no consciousness and, in my opinion, no intelligence (this of course
is just Dreyfus' and Searle's arguments).
I think this discussion would profit if "self-reproduction" were
defined. Cugini is correct, I'd say, in pointing out that self-
reference does not equal intelligence or consciousness (i.e., the
Hofstadter argument pushed to the extreme). I think that we ought
to consider self-reference as "self-communication" or "symbolic
self-interaction." See Dennett's great book, Elbow Room. A computer
has self-reference, but no self-communication: A computer can "say"
"I'm a computer" (self-reference), but that "utterance" doesn't
mean anything to the machine (no self-communication); again, this
is just Searle's argument put in terms of Dennett. "Self-communication"
is important because it changes our ideas about other human
cognitive processes and the whole size/self-reference problem.
Memory is not a data-structure, but a way that a human communicates
with himself/herself through time (Michael Cole's argument). That's
why it doesn't matter how large the memory is: if memory
is not a self-communication method, the organism will never
be conscious. In any case, Dennett's book is excellent on this
problem, as is Searle's Minds, Brains and Science. Some of these
issues are addressed directly in the journal Human Development,
though that journal may be a little too social-psychological for
the hardcore AI community.
------------------------------
End of AIList Digest
********************