LAWS@SRI-AI.ARPA (03/15/85)
From: AIList Moderator Kenneth Laws <AIList-REQUEST@SRI-AI.ARPA>
AIList Digest Friday, 15 Mar 1985 Volume 3 : Issue 34
Today's Topics:
AI Sites - Universities,
Linguistics - Hangul,
Literature - The Journal for the Integrated Study of Artificial
Intelligence, Cognitive Science and Applied Epistemology,
Policy - Humor,
Humor - AI Joke Contest & Eliza & Political/Mathematical Humor & ))) &
Greedy Algorithms,
Course - Nonexistent Objects and Fiction (CSLI)
Seminars - Knowledge-Based Software Development at Kestrel (SU) &
Computational Geometry, Rewrite Rules, etc. (PARC) &
Shape from Function (MIT)
----------------------------------------------------------------------
Date: Thursday, 14 Mar 85 12:55:07 EST
From: jaffe (elliot jaffe) @ cmu-psy-a
Subject: Universities and AI
About 4 months ago on AILIST I remember a discussion about which universities
were doing AI work and a general listing was given.
1) Can I get a copy of that set of lists?
2) Does anybody have any new information to enter?
Elliot Jaffe
Jaffe@CMU-PSY-A.ARPA
[I'll send Elliot a copy of the list of sites receiving AIList.
See also the November 1983 issue of IEEE Spectrum; it misses
some sites -- SRI, for example -- but the chart of academic and
nonacademic sites on pp. 59-68 is impressive. -- KIL]
------------------------------
Date: 14 Mar 85 08:14 PST
From: Kay.pa@XEROX.ARPA
Subject: Hangul
The word "Hangul" (not "Hungal") may refer to a dialect, but that is not
its main use. It is the name of the wonderfully ingenious writing
system that Korean has, named after the emperor who invented it. The
system is interpretable both phonemically and syllabically and is a gem
of design. There are the same number of good automated translation
systems for this as for every other natural language, namely none.
--Martin Kay.
[The last IEEE Computer featured oriental text entry systems; I
remember seeing Korean discussed, so perhaps there are some
appropriate references. -- KIL]
------------------------------
Date: Wed, 13 Mar 85 08:31:31 pst
From: Vaughan Pratt <pratt@Navajo>
Subject: A place for everything
[Forwarded from the Stanford bboard by Laws@SRI-AI.]
I just received volume 1, number 1 of The Journal for the Integrated
Study of Artificial Intelligence, Cognitive Science and Applied Epistemology.
It is described as "a quarterly journal published by Communication and
Cognition-AI at the State University of Ghent, Blandijnberg 2,
B-9000 Ghent, Belgium."
I imagine they will soon tire of allusions to Browning's "How They Brought
the Good News from Ghent to Aix," if they haven't already.
-v
------------------------------
Date: 14 Mar 85 15:30:43 EST
From: Tim <WEINRICH@RUTGERS.ARPA>
Subject: Policy on Humor
[...The policy of this list is to avoid "ethnic" humor or any form of
wit at the expense of individuals or sensitive groups. There may
have been a few lapses, but I'll try to enforce the policy strictly
in the future. -- KIL]
Unfortunately, if you really intend to stick strictly to this policy,
I'm afraid you will end up posting no humor at all, as it seems there are
very few statements and even fewer jokes which no group finds offensive.
Do you realize that, according to this policy, you cannot publish any of
the various humorous versions of Little Red Riding Hood?
I mourn the demise of humor on AIList.
Twinerik
------------------------------
Date: Wed, 13 Mar 85 16:31:33 pst
From: newton@cit-vax (Mike Newton)
Subject: rape and humor
Though rape is a barbaric act, censorship is a far worse crime for it
affects so many more. Do the people who object so much about the "polly
nomial" story complain so loudly w.r.t. jokes about murder, or, worse,
war?
There is a fundamental difference between an act and writing about an act.
If jokes about a subject are banned, how long is it before satire about
the same subject is banned? Then there is only a small step fiction is
banned. Next comes fact, research...
mike
ps: I realize that the above is a viewpoint some will find objectionable.
Arguments about my first sentence had better be convincing: Though due to
my sex it would be hard to rape me, I have been shot, mugged and robbed --
one of these incidents still bothers me noticeably -- yet I do NOT think
that jokes about these subjects should be censored.
------------------------------
Date: Thu 14 Mar 85 11:11:57-PST
From: Wilkins <WILKINS@SRI-AI.ARPA>
Subject: in defense of humor
I am all for decent people expressing opinions and running the
government, but being oversensitive can take the fun out of life.
Any reasonable definition of rape should include the fact that
the raper and rapee are human or at least members of the
animal kingdom. I find it hard to be offended when polynomials
are raped, however. We all know mathematical/computer language
is colorful, but using a verb that would be in poor taste when
applied to humans does not mean its in poor taste when applied
to abstract data structures. For example, killing a line in
your editor is not considered murder.
While the story in question did say Polly Nomial was a
"paragon of womanly virtue", it never said she was a woman
or was human, so one can easily assume she is a polynomial.
The story implies that Curly Pi is a "common fraction" which
is consistent with these characters being abstract mathematical
being. This story was good humor, though admittedly non-mathematicians
might not appreciate it and transfer it into a human domain and
be offended. I personally enjoyed getting the chance to read it.
Striking a blow for more humor and fun in life,
David
------------------------------
Date: Tue 12 Mar 85 13:26:45-EST
From: Bob Hall <RJH%MIT-OZ@MIT-MC.ARPA>
Subject: AI Jokes RESULTS
[Forwarded from the MIT bboard by SASW@MIT-MC.]
Q. "What do you get when you cross an expert system with an
orangutan?"
A. "Another Harry Reasoner."
Yes folks, that's it, the winner of the first annual AI Jokes Contest.
The winner has been awarded his lovely blue and gold Cal T-shirt.
We got many entries (single digits, though), which were much appreciated.
Highlights follow. The names have been withheld just in case the authors
are sensitive about being associated with their entries. (Anyone who
wants explicit credit should send me mail.)
Consensus honorable mentions:
Q. "What did the Apple 5e say when a human fell on it?"
A. "Ureka! (sic) I've discovered gravity!"
Q. "How many expert systems does it take to screw in a light bulb?"
A. "If they're so smart, why don't they know?"
*******************************************************************
Now that you've got a feel for the competition, ANNOUNCING
AI JOKES II: THE WRATH OF CONS
Send your entry for the next, bigger, better AI Jokes Contest to
either me (rjh%mit-oz@mit-mc) or US mail directly to
AI JOKES II: THE WRATH OF CONS
1717 Allston Way
Berkeley, CA 94703
Terms are the same: if yours is the funniest, as judged by an impartial
panel of natural intelligences, you win your choice of a T-shirt from
one of the following schools:
U.C. Berkeley
Cal State, Hayward
Chabot College, Hayward
MIT
Harvard
Leland Stanford Junior College
Doug Flutie's alma mater
Enter Now!! (Include shirt size and school preference)
*******************************************************************
------------------------------
Date: 14 March 1985 0512-EST
From: Jeff Shrager@CMU-CS-A
Subject: Security the MIT way
[Forwarded from the CMU bboard by Laws@SRI-AI.]
If you have access to MIT-OZ, try sending a message to their operator
from a not-logged-in task. In order to handle dial-in randoms who tend
to do operator sends (and used to get no reply) they've now piped the
operator's message buffer through an Eliza system, which proceeds to
"analyze" the luser's problems. Hack Hack.
------------------------------
Date: Wed 6 Mar 85 08:57:18-PST
From: Ken Laws <Laws@SRI-AI.ARPA>
Subject: Political/Mathematical Humor
From a WORKS message by Greg Kuperberg (harvard!talcott!gjk or
talcott!gjk@topaz):
"2*x^5-10*x+5=0 is not solvable by radicals." -Evariste Galois.
------------------------------
Date: Sun 24 Feb 85 14:13:34-PST
From: Steven Tepper <greep@SU-CSLI.ARPA>
Subject: Found: right parens
[Forwarded from the Stanford AI bboard by Laws@SRI-AI.]
I found a sequence of right parentheses lying around on the Ethernet.
They look like they might have fallen off the end of somebody's Lisp program.
If you lost some and think these might be yours, send me a message identifying
them.
Also, has anyone seen a matched pair of asterisks in Gacha 10? I can get
new ones, but this particular pair had sentimental value.
------------------------------
Date: Sun 24 Feb 85 15:31:58-PST
From: Gustavo Fernandez <FERNANDEZ@SU-SCORE.ARPA>
Subject: RE: Found: right parens
[Forwarded from the Stanford AI bboard by Laws@SRI-AI.]
Sorry, I have not seen your asterisks, but please! Does anyone know where
the daily bit buckets are stored? I lost some data Thusday night by
doing one too many left shift on a program I was writing on SCORE and
I was wondering whether I might be in some way able to recover the bit
strings. Any help would be greatly appreciated.
Gus Fernandez
FERNANDEZ@SCORE
------------------------------
Date: Thu, 7 Mar 85 17:57:17 est
From: Walter Hamscher <hamscher at mit-htvax>
Subject: Graduate Student Lunch *SEMINAR*
[Forwarded from the MIT bboard by SASW@MIT-MC.]
Place: *THIRD FLOOR THEORY PLAYROOM*
Date: Friday, March 8
Time: 12 Noon
Hosts: Isaac Kohane and Mike Wellman
COMPUTER AIDED CONCEPTUAL ART
REVOLTING SEMINAR SERIES
presents
GREEDY ALGORITHMS FOR REALLY HARD PROBLEMS
Mike Generous
Penny Weise
Dahlia Fulisch
Perhaps the best known greedy algorithm is Kruskal's Minimum
Spanning Tree algorithm. The concept of "greedy algorithm"
is here generalized to provide a framework in which any
problem can be solved with bounded error in constant time.
The simplest such algorithm is a stupefyingly
straightforward algorithm for finding the maximum of a list
of numbers: take the maximum of the first two elements of
the list. The bounded error arises from the provability of
the correctness criterion "We could have done worse!". The
Satisfiability problem for the predicate calculus is solved
trivially, because we have shown conclusively elsewhere that
we are Satisfied with predicate calculus. The framework
applies also to meta-problem solving (i.e. "What problem
should we sic the greedy algorithm on next"), as we will
demonstrate by applying it to the problem of generating
grant proposals. If there is time, we will give other
examples involving Single-State Automata (Moronotrons) and
Extremist Graph Theory.
------------------------------
Date: Thu 14 Mar 85 16:42:26-PST
From: Emma Pease <Emma@SU-CSLI.ARPA>
Subject: Course - Nonexistent Objects and Fiction (CSLI)
[Excerpted from the CSLI Newsletter by Laws@SRI-AI.]
COURSE ANNOUNCEMENT
Graduate Seminar - Philosophy
``Nonexistent Objects and the Semantics of Fiction''
Edward N. Zalta, CSLI
The problem of how it is we can think about and tell stories about
what does not exist is one of the foremost problems in the study of
intentionality. We'll begin by asking what an analysis of fiction,
and stories in general, ought to do, and then quickly review the
problems facing the semantic analysis of sentences about nonexistent
objects developed by Meinong, Russell, Quine, and the free logicians.
We then turn to a careful presentation of both Terence Parsons'
neo-Meinongian views (developed in his book: Nonexistent Objects) and
my own, which has a Meinongian flavor. There will be a comparison of
how the language and logic of these theories represent the meaning of
English sentences about nonexistents. Then we shall ask whether these
theories provide a better representation, and do a better job of
analyzing fiction in general, than some current alternatives, some of
which do without nonexistents (Plantinga, Searle, Fine, Lewis) and
some of which appeal to some sort of abstract objects (Kripke, van
Inwagen, Wolterstorff). We'll conclude the course with a brief
examination of how these axiomatized theories fit into a larger
picture of the semantics of language and intensionality.
The first meeting of this seminar will be held in the Venture Hall
trailers conference room, Tuesday April 2, at 1:15.
------------------------------
Date: Mon 11 Mar 85 17:04:08-PST
From: Paula Edmisten <Edmisten@SUMEX-AIM.ARPA>
Subject: Seminar - Knowledge-Based Software Development at Kestrel (SU)
[Forwarded from the Stanford SIGLUNCH distribution by Laws@SRI-AI.]
DATE: Friday, March 15, 1985
LOCATION: Chemistry Gazebo, between Physical and Organic Chemistry
TIME: 12:05
SPEAKER: Douglas R. Smith, Kestrel Institute, Palo Alto, CA.
ABSTRACT: KNOWLEDGE-BASED SOFTWARE DEVELOPMENT AT KESTREL INSTITUTE
Kestrel Institute is a non-profit organization with a two-fold
purpose:
Research and graduate education in computer science.
Its main research goal is the formalization and incremental
automation of software development.
Towards this goal, we carry out research in such areas as machine
intelligence, very-high-level languages, algorithm design,
transformation and synthesis, software project management, and
knowledge-base programming environments.
I'll present an overview of current and planned research at Kestrel
Institute and describe our experience with the DSE system, a
knowledge-base programming environment in routine use. The bulk of
the talk will examine two research themes at Kestrel:
Knowledge compilation and,
The use of schemes in program synthesis.
Knowledge compilation involves transforming declarative knowledge plus
directions for its useage into efficient procedural form. The uses of
program schemes and strategies for instantiating them include
knowledge compilation and the design of algorithms from
specifications.
Paula
------------------------------
Date: 12 Mar 85 15:53:13 PST
From: yao.pa@XEROX.ARPA
Subject: Seminars - Computational Geometry, Rewrite Rules, Etc. (PARC)
[Excerpted from the Stanford bboard by Laws@SRI-AI.]
The next BATS [Bay Area Theory Seminar] will take place on Friday,
March 15 at Xerox PARC (in the auditorium). [...] -- Frances
SCHEDULE
10 a.m. Ronald Graham (AT&T Bell Lab): Remarks on The Finite Radon
Transform
11 a.m. Frances Yao (Xerox PARC): A General Approach to d-Dimensional
Geometric Queries
1 p.m. Andrew Yao (Stanford): Separating the Polynomial-Time
Hierarchy by Oracles
2 p.m. Joe Halpern (IBM): What Does It Mean for Rewrite Rules to be
"Correct"?
3 p.m. Andrei Broder (DEC): Vote Early and Vote Often; The
Distributed Lottery Problem
[...]
A General Approach to d-Dimensional Geometric Queries
Frances Yao
Xerox Palo Alto Research Center
ABSTRACT Research results in the area of computational geometry have
been largely limited to problems concerning simple relations between
objects in the plane. In this talk, we shall define what we call a
"generic query" in d-space, and present a uniform solution to it. As
examples of application, the nearest-neighbor query in d-space can be
solved in linear space and sublinear time; the pairwise intersection
problem for polytopes and the construction of minimum spanning trees in
d-space can be solved in linear space and subquadratic time.
What does it mean for rewrite rules to be "correct"?
Joe Halpern
IBM San Jose Research Center
ABSTRACT We consider an operational definition for FP via rewrite
rules. What would it mean for such a definition to be correct? We
certainly want the rewrite rules to capture correctly our intuitions
regarding the meaning of the primitive functions. We also want there to
be enough rewrite rules to compute the correct meaning of all
expressions, but not too many, thus making equivalent two expressions
that should be different. And what does it mean for there to be
"enough" rules?
We give a formal criterion for deciding whether there are enough rewrite
rules and show that our rewrite rules meet that criterion. We develop
powerful techniques to prove these results, that
involve imposing a notion of types on the untyped language FP, and then
using techniques of typed lambda-calculus theory. (Note: This talk is
completely self-contained. No previous knowledge of FP or lambda
calcuclus will be assumed.)
------------------------------
Date: 12 Mar 1985 22:11 EST (Tue)
From: "Daniel S. Weld" <WELD%MIT-OZ@MIT-MC.ARPA>
Subject: Seminar - Shape from Function (MIT)
[Forwarded from the MIT bboard by SASW@MIT-MC.]
AI Revolving Seminar Tues 3/19/85 4:00pm 8th floor playroom
SHAPE FROM FUNCTION VIA MOTION ANALYSIS
with Application to the Automatic Design of
Orienting Devices for Vibratory Part Feeders
TOMAS LOZANO-PEREZ
This talk explores the premise that a device's function can be
characterized by how it interacts with other objects. I focus on
devices for which some aspect of their function can be characterized
in terms of constraints they place on the motions of objects.
Commonplace examples of this class of device abound; any office is
full of them: chairs, cups, rulers, telephones, cabinets, bookends,
etc. In fact, the shape of most objects is constrained by the legal
motions of that object and of objects it must interact with.
I suggest a representation for the function of this class of objects
in terms of motion contraints. These possible--motion constraints are
expressed as an abstract diagram. Combinations of these diagrams
serve both in describing a device's function and in designing devices
with specified behavior.
The design problem for these devices is a kind of inverse of the
motion planning problem in robotics. In both cases we know the shape
of the moving part. In motion planning, we are given the obstacles
and we must find a legal path between the specified origin and
destination. In our view of design, however, we are given the desired
motion (actually a range of possible motions) and are asked to find a
legal shape of the obstacle, that is, the device.
We illustrate our approach to design with a detailed case study of
mechanical part feeders, a class of real devices with an interesting
and direct relationship between shape and function.
------------------------------
End of AIList Digest
********************