Copy Link
Add to Bookmark
Report
AIList Digest Volume 2 Issue 047
AIList Digest Sunday, 15 Apr 1984 Volume 2 : Issue 47
Today's Topics:
Education - Request for Writing, Geometry Systems,
Msc. - TAMITTC,
Applications - Artificial Big Brother,
Seminars - Qualitative Process Theory & AI and VLSI,
Conferences - COLING84 Information
----------------------------------------------------------------------
Date: Sat, 14 Apr 84 02:06 EST
From: Malcolm Cook <cook%umass-cs.csnet@csnet-relay.arpa>
Subject: TEACHING SYSTEMS. WRITING & GEOMETRY.
Im looking for information on 2 things:
1) what systems exist that are used to teach language skills,
especially teaching writing to children. also, I remember hearing
about studies showing that children are motivated to write simply
by having a simplified editor. Any pointers to this study(s)?
There was an interesting article in N.Y. times sunday mag section
2-3 months back entitles "Writing to Read", and was about a curriculum
for 1st & 2nd grade, in which the children were learning a simple
morpheme<==>grapheme map, allowing them to pphonetically spell
any word they could pronounce. Are there any AI systems involved
in this course?
2) What does there exist for tutoring sytems in geometry.
I am somewhat familiar with the CMU approach (Boyle, Anderson, Shrager...)
but what else is around? Anything on the spectrum from GOOD
programmed instruction to /reactive environments/ would be of interest.
thanks,
Malcolm Cook (Cook.umass-cs@csnet-relay)
------------------------------
Date: 7 Apr 84 17:39:57-PST (Sat)
From: decvax!cwruecmp!borgia @ Ucb-Vax
Subject: TAMITTC
Article-I.D.: cwruecmp.1135
A few weeks ago a cryptic TAMITTC poster appeared on all the doors
of our computer engineering department. Recently I discovered what
it was all about on an obscure campus billboard.
There Are More Important Things Than Computers
And there was a footnote.
Like what? People? Oh, those things!
------------------------------
Date: 8 Apr 84 15:53:05-PST (Sun)
From: harpo!ulysses!allegra!don @ Ucb-Vax
Subject: Artificial Big Brother
Article-I.D.: allegra.2388
AI and criminology
I totally agree with DW's article in net.crypt. Computer techniques
would be used to keep track of "political criminals". Middle class
intellectuals are far more vulnerable to this sort of control than are
street criminals and drifters.
Already, right wing organizations use this technology to keep track of
people they consider politically dangerous, and while the government is
not allowed to do this, they have received information from these
organizations under the table. In some cases, victims are chosen
simply by correlating magazine subscription information.
------------------------------
Date: 9 Apr 84 10:05:30-PST (Mon)
From: hplabs!hao!seismo!brl-vgr!abc @ Ucb-Vax
Subject: Re: Artificial Big Brother
Article-I.D.: brl-vgr.15
But just like so many things, do you really think that,
because the "good guys" don't build a tool which can be
used for "good" or "evil" that the "bad guys" won't build
and use the tool?
It seems that what is needed is research into methods for
controlling these tools (Computer Science) and research
into new public policies regarding the use and misuse of
such tools (Humanities and Social Science).
Remember: whether the U.S. did it or not, others still
would have developed and deployed nuclear weapons.
------------------------------
Date: 10 Apr 84 13:51:13-PST (Tue)
From: hplabs!tektronix!tekigm!dand @ Ucb-Vax
Subject: Re: Artificial Big Brother
Article-I.D.: tekigm.74
I cannot agree too strongly with Brint Cooper about this. The tool never makes
the wielder any more or less an "evil" person.
If given a choice, I'd rather have such a tool built by the established AI
community for two reasons:
1) The program's existance is published, so people can think about the
implications and possibly set up systems to reduce the amount of abuse
the system is used for. As a possible victim of misuse, I can also start
thinking about preventive measures to unreasonable privacy invasions(I
personally believe no one even now has any real privacy if someone is out
to do you in, but that is not germain here.)
2) If such a tool is in the public domain, at least the people it was
originally designed for, the law-enforcement agencies, would get some
use out of it. If this tracking system were to be built in a CIA shop or
an NSA shop, no one outside those organizations will ever know of its
existance, and thus never be able to use it.
Abuses with such a system are going to be inevitable; the goal for us to set
is to see that the abuses are kept to a minimum, which we can't do if the
system requires "Top Secret/Burn Before Reading" clearance to even know that it
exists.
Lest anyone try to say that the possible abuses of such a system outweigh the
few benefits of it, remember that Theodore Bundy was convincted with such
evidence as gasoline receipts in the area where one of his victims disappeared,
at the same time she disappeared. With such a system, perhaps, Ted Bundy would
not have racked up the score of dead, young women that he did. Such a system
might help pinpoint the current Green River Killer in Washington State,
or reduce the predations of the itinerant killers who prey on anyone they
think they can get away with. If some shadowy bureaucrat were out to get you,
this system would not be necessary--a judge's signature is all that is needed
to open up the records of your Visa, your bank, your employer, etc (granted it
may not be a legal action on that part of that judge, but we're already talking
about illegal activities, no?).
Finally, if this discussion is going to go on, let's move it to net.politics
or net.misc or net.legal, net.ai is not the proper forum for this discussion.
Dan C Duval
ISI Engineering
Tektronix,Inc
tektronix!tekigm!dand
------------------------------
Date: 12 Apr 1984 16:16 EST (Thu)
From: Cobb%MIT-OZ@MIT-MC.ARPA
Subject: Qualitative Process Theory
[Forwarded from the MIT bboard by SASW@MIT-MC.]
SEMINAR
Kenneth D. Forbus
April 17 - 4:00PM
NE43 - 8th floor Playroom
Title: QUALITATIVE PROCESS THEORY
Objects move, collide, flow, bend, heat up, cool down,
stretch, break, and boil. These and other things that happen to cause
changes in objects over time are intuitively characterized as
processes. To understand common sense physical reasoning and make
programs that interact with the physical world as well as people do we
must understand qualitative reasoning about processes and their
effects. Qualitative Process theory defines a simple notion of
physical process that appears useful as a language in which to write
dynamical theories. Reasoning about processes also motivates a new
qualitative representation for quantity in terms of inequalities,
called the Quantity Space.
This talk will describe the basic concepts of Qualitative
Process theory, two different kinds of reasoning that can be performed
with them, and its implications for causal reasoning. Several
examples will be presented to illustrate the utility of the theory,
including figuring out that a boiler can blow up and how different
theories of motion may be encoded.
Refreshments at 3:45PM
Host: Professor Patrick H. Winston
------------------------------
Date: 12 Apr 84 16:31:46 EST
From: Smadar <KEDAR-CABELLI@RUTGERS.ARPA>
Subject: III Seminar on AI and VLSI this Coming Thursday (room 423)...
[Forwarded from the Rutgers bboard by Laws@SRI-AI.]
I I I SEMINAR
Title: Knowledge-Based Aids for VLSI Design
Speaker: Tom Mitchell
Date: Thursday, April 19, 1984, 1:30-2:30 PM
Location: Hill Center, ***room 423***
Knowledge-based systems provide one possible approach to dealing
with the complexities of VLSI design. This talk discusses the design
of such a system, called VEXED, which aids the user in the interactive
design of VLSI circuits. VEXED is intended to provide suggestions
regarding alternative implementations of circuit modules, as well as
warnings regarding conflicting constraints during design. The system
is composed of a circuit network expert (CIRED), a layout expert (RPL),
and a digital signal analysis expert (CRITTER). A prototype version of
VEXED has been implemented, and a second version of the system is now
under development.
------------------------------
Date: Thu 12 Apr 84 11:10:48-PST
From: Don Walker <WALKER@SRI-AI.ARPA>
Subject: COLING84 information on registration, travel, housing,
summer school
****************** PLEASE POST, CIRCULATE, AND REDISTRIBUTE *****************
COLING84, TENTH INTERNATIONAL CONFERENCE ON COMPUTATIONAL LINGUISTICS
COLING84 is scheduled for 2-6 July 1984 at Stanford University, Stanford,
California. It will also constitute the 22nd Annual Meeting of the
Association for Computational Linguistics, which will host the conference.
Information about the conference, registration, travel, and accommodations,
and about the six summer school courses that will be held during the
preceding week (25-29 June) has just been made available in the form of The
COLING Courier. For a copy, contact Don Walker, COLING84, SRI
International, Menlo Park, California, 94025, USA [phone: (415)859-3071;
arpanet: walker@sri-ai; telex [650] 334486]. Other requests for information
about the conference should be addressed to Martin Kay, COLING84, Xerox PARC,
3333 Coyote Hill Road, Palo Alto, California 94304, USA [phone:
1-(415)494-4428; arpanet: kay@xerox; telex [650] 1715596].
The summer school, which will be held 25-29 June, consists of week-long
tutorials on six subjects that are central to computational linguistics but
on which instruction is still not routinely available: LISP AS
LANGUAGE--Brian Smith, Xerox and Stanford; PROLOG FOR NATURAL LANGUAGE
ANALYSIS--Fernando Pereira, SRI International; PARSER CONSTRUCTION
TECHNIQUES--Henry Thompson, Edinburgh; SITUATION SEMANTICS--David Israel,
BBN, & John Perry, Stanford; MACHINE TRANSLATION--Brian Harris, Ottawa, &
Alan Melby, Brigham Young; SOUND STRUCTURE OF LANGUAGE--Mark Liberman, Bell
Labs. Enrollments are limited to 30 in each tutorial, so register early.
A remarkably rich set of computational facilities will be available at
Coling84 for demonstrating programs and systems. For information, contact
Doug Appelt, SRI International, Menlo Park, California 94025 [phone: (415)
859-6150; arpanet: appelt@sri-ai; telex: [650] 334486].
You are advised to BOOK EARLY FOR COLING84, since airline reservations will
be much harder than usual to obtain. Custom Travel Consultants, 2105
Woodside Road, Woodside, CA 94062 [phone (415)369-2105], is responsible for
registration, travel, and housing. Full information is provided in the
Coling Courier, but call them if time is short.
------------------------------
End of AIList Digest
********************
