Copy Link
Add to Bookmark
Report
AIList Digest Volume 2 Issue 050
AIList Digest Saturday, 21 Apr 1984 Volume 2 : Issue 50
Today's Topics:
AI Programming - Characterization & Software Engineering,
AI Literature - Computer Database & Metaphor and Sociolinguistics &
Automated Reasoning Book,
Expert Systems - DARPA Sets Expert System Goals,
Administrivia - Creation of Pascal Mailing List,
Humor - Lady Lovelace's Encryption Algorithm,
Seminars - Model-Based Vision & Robot Design Issues
----------------------------------------------------------------------
Date: Mon 16 Apr 84 13:52:04-PST
From: Tom Dietterich <DIETTERICH@SUMEX-AIM.ARPA>
Subject: RE: AI Programming
I think a major difference between most AI programming and most non-AI
programming is that AI programming usually involves implementing
additional layers of interpretation on top of whatever programming
system is being employed. Any system that needs to reason about its
own actions, its own assumptions, and so on, requires this extra layer
of interpretation. The kinds of programs that I work on--learning
programs--also need to modify themselves as they run. This helps
explain why LISP is so popular--it provides very good support for
building your own interpreters: the ability to dynamically define new
symbols, the ability to construct arbitrary binding environments, and
the ability to invoke EVAL on arbitrary expressions. Perhaps LISP
is best viewed as a interpreter language rather than a programming
language.
--Tom
------------------------------
Date: 16 Apr 84 6:18:16-PST (Mon)
From: hplabs!hao!seismo!rochester!ritcv!ccivax!band @ Ucb-Vax
Subject: Incomplete specifications ...
Article-I.D.: ccivax.111
In reference to the recent discussion about Software
Engineering and incomplete specifications.
For any new computer system, specifications at some
point must be incomplete. A computer program is a
new machine -- it's never been constructed before.
So the final details always remain until the end. This does
not mean that one does not begin construction. On the
contrary, it seems to this writer that all too often
construction begins before any specifications are written.
What's needed is a middle path. Design people need
enough requirements and constraints ( specifications )
to start work. What should be provided is concise
documentation of the requirements and constraints, as
well as documentation of the unknowns and the risk.
Designers as they work will learn more about what is
and is not possible and this information will refine
the specifications. But holes will remain. This kind
of "evolutionary" development has been described by
Carl Hewitt in an article entitled "Evolutionary
Programming" in SOFTWARE ENGINEERING, edited by
H. Freeman and P.M. Lewis II (NY: Academic Press, 1980).
I submit that any computer system development must
be a risk, and that it can only be developed by proceeding
with incomplete specifications. The complement to this
is that large projects must be reviewed for viability
as knowledge is gained through this evolutionary
growth. Sometimes it's better to quit before good money
is wasted.
There's more to this issue that what is written here.
But it is not correct to hold AI programming up as some
sort of magical paradigm that is not subject to rudimentary
engineering discipline. Software Engineering may indeed
have much to learn from the AI style of programming,
but programming in general has much to learn from engineering
disciplines also.
Bill Anderson
...!ucbvax!amd70!rocksvax!ritcv!ccivax!band
...!{allegra | decvax}!rochester!ccivax!band
------------------------------
Date: Tue 17 Apr 84 09:31:17-PST
From: C.S./Math Library <LIBRARY@SU-SCORE.ARPA>
Subject: New File on Dialog-- The Computer Database
[Forward from the Stanford bboard by Laws@SRI-AI.]
The Computer Database is a new file on Dialog which covers computers,
telecommunications and Electronics. The file went online in January and
covers material from 1983 to date. The documentation which comes with
the file has a thesaurus which appears to be very up to date in terminlology
for online searching. The journals indexed include ACM publications, AI,
Industrial Robot, SIAM publications, IEEEE as well as Infoworld, PC World,
Dr. Dobbs, Byte etc.
[...]
Harry
------------------------------
Date: Wed, 18 Apr 1984 13:18:02 EST
From: FF Bottles of Beer on the Wall,...
<AXLER%upenn-1100.csnet@csnet-relay.arpa>
Subject: Discounted Books
A number of books on metaphor and sociolinguistics that I mentioned
in an earlier message are now on sale by their publisher, the University of
Pennsylvania Press. The sale catalog is available by writing them at
3933 Walnut Street, Philadelphia, PA 19104. Minimum order is $10.00.
Among the items available and of interest to AI researchers are:
Sapir & Crocker, "The Social Use of Metaphor" $8.75 (50% off)
Hymes, "Foundations in Sociolinguistics" $6.97 (30% off)
Kirschenblatt-Gimblett, "Speech Play", $6.00 (70% off)
Weinreich, "On Semantics" $10.50 (70% off)
Labov, "Sociolinguistic Patterns", $10.00 (60% off)
Maranda & Maranda, "Structural Analysis of Oral Tradition", $8.40 (60%off)
--Dave Axler
------------------------------
Date: 28-Mar-84 12:33:59-CST (Wed)
From: Larry Wos <Wos@ANL-MCS>
Subject: Automated Reasoning
[Forwarded from the Prolog Digest by Laws@SRI-AI.]
The book, Automated Reasoning: Introduction and Applications, by
Wos, Overbeek, Lusk, and Boyle, is now available from Prentice-
Hall. It introduces basic concepts by showing how an automated
reasoning program can be used to solve various puzzles. The puz-
zles include the "truthtellers and liars" puzzle that was exten-
sively discussed in the Prolog Digest, McCarthy's domino and
checkerboard puzzle, and the billiard ball and balance scale puz-
zle. The book is written in a somewhat informal style and no
background is required. It also contains a rigorous treatment of
the elements of automated reasoning. The book relies heavily on
examples, includes many exercises, and discusses various applica-
tions of automated reasoning. The applications include logic
circuit design, circuit validation, research in mathematics,
research in formal logic, control systems, and program verifica-
tion. Other chapters of the book provide an introduction to Pro-
log and to expert systems. The last chapter, "The Art of Au-
tomated Reasoning", gives guidelines for choosing representation,
inference rules, and strategies.
The book is based on examples actually solved by existing
automated reasoning programs. Certain of these programs are
available and portable. The book can be used as a college text,
consulted by those who wish to study possible applications, or
simply read by the curious.
It can be ordered directly from Prentice-Hall with a visa
or master card by calling 800-526-0485 and the ISBN number
is 0-13-054446-9 for the soft cover. The soft cover
is 18.95, and the hard 28.95.
-- LW
------------------------------
Date: 17-Apr-84 17:24 PST
From: William Daul OAD / TYMSHARE / McDonnell Douglas
<WBD.TYM@OFFICE-2.ARPA>
Subject: DARPA Sets Expert System Goals
From DEFENSE ELECTRONICS (April 1984):
Among the goals established for DARPA's expert systems technology program are
increased storage capacity and reasoning power that can deal with 10,000 rules
and provide 4,000 rule inferences per second for stand-alone systems and 30,000
rules and 12,000 inferences per second for multiple cooperating expert systems.
The program, part of DARPA's strategic computing initiative, is aimed at
achieving a framework to support battle management applications. The Air
Force's Rome Air Development Center will be issuing RFPs in nine technical
areas: explanation and presentation capability, ability to handle uncertain and
missing knowledge, fusion of information from several sources, flexible control
mechanisms, knowledge acquisition and representation, expansion of knowledge
capacity and extent, enhanced inference capability, exploiting expert systems
on multiprocessor architectures, and development of cooperative distributed
expert systems. Multiple contract awards are planned for each area, and one
or two additional awards are planned for complete system development.
------------------------------
Date: Wed, 11 Apr 84 8:48:51 EST
From: "Ferd Brundick (VLD/LTTB)" <fsbrn@Brl-Voc.ARPA>
Subject: Creation of new mailing list
Hi,
A new special interest mailing list called info-pascal has been
created. Enclosed below is the summary for the list. If you would
like to be added to the list, please check with your local Postmaster
or send a message to info-pascal-request@brl-voc.
dsw, fferd
Fred S. Brundick
aka Pascal Postman
USABRL, APG, MD.
<info-pascal-request@brl-voc>
-----------------------------------------------------------
INFO-PASCAL@BRL-VOC.ARPA
This list is intended for people who are interested in the programming
languages Pascal and Modula-2. Discussions of any Pascal/Modula-2 imple-
mentation (from mainframe to micro) are welcome.
Archives are kept on SIMTEL20 in the files:
MICRO:<CPM.ARCHIVES>PASCAL-ARCHIV.TXT (current archives)
MICRO:<CPM.ARCHIVES>PASCAL.ARCHIV.ymmdd (older archives)
All requests to be added to or deleted from this list, problems, questions,
etc., should be sent to INFO-PASCAL-REQUEST@BRL-VOC.ARPA.
Coordinator: Frederick S. Brundick <fsbrn@brl-voc.arpa>
------------------------------
Date: 19 Apr 1984 12:34:36-EST
From: walter at mit-htvax
Subject: Seminar - Lady Lovelace's Encryption Algorithm
[Forwarded from the MIT bboard by SASW@MIT-MC.]
ANNALS OF COMPUTER SCIENCE SEMINAR
DATE: Friday, April 20th, 1984
TIME: Refreshments 12:00 noon
PLACE: MIT AI Lab 8th Floor Playroom
LADY LOVELACE'S ENCRYPTION ALGORITHM
ABSTRACT
Znk loxyz iusv{zkx vxumxgsskx }gy g totkzkktzn3iktz{x tuhrk}usgt2
Rgj G{m{yzg Gjg Hxut Ru|krgik2 jg{mnzkx ul znk vukz Ruxj Hxut.
Gy g zkktgmkx2 G{m{yzg joyvrgkj gyzutoynotm vxu}kyy ot sgznksgzoiy.
]nkt ynk }gy komnzkkt g{m{yzg loxyz yg} Ingxrky Hghhgmk-y gtgrzoigr
ktmotk2 g igri{rgzotm sginotk zngz }gy znk luxkx{ttkx ul znk sujkxt
iusv{zkx. Ot komnzkkt luxz3z}u2 ynk zxgtyrgzkj g vgvkx ut znk
ktmotk lxus Lxktin zu Ktmroyn gjjotm nkx u}t |ur{sotu{y tuzky. Ot
y{hykw{ktz }xozotmy ynk jkyixohkj znk (ruuv( gtj (y{hxu{zotk(
iutikvzy g iktz{x hkluxk znkox osvrksktzgzout ot krkizxutoi
jomozgr iusv{zkxy .h{z gy lgx gy O qtu}2 nu}k|kx2 ynk tk|kx joj
gtznotm }ozn ktixvzout/. Rgj Ru|krgik gtj Hghhgmk ngj g rutm
gtj iruyk lxoktjynov gtj ynk }gy g jkjoigzkj vgxztkx ot noy }uxq
}ozn znk gtgrzoigr ktmotk. [tluxz{tgzkr ynk }gy nkrj hgiq h
gtzo3lksotoyz gzzoz{jky gtj h nkx u}t uhykyyout }ozn mgshrotm ut
nuxyk xgiky. Rgj Ru|krgik jokj ul igtikx gz gmk znoxz3yo~. Tu}
zngz u{|k jkiujkj znoy skyygmk2 rkz-y grr mkz hgiq zu }uxq.
This fascinating historical discussion and the
accompanying Graduate Student Lunch will be hosted
by Dan Carnese and Maria Gruenewald.
------------------------------
Date: 18 Apr 1984 14:38 EST (Wed)
From: Cobb%MIT-OZ@MIT-MC.ARPA
Subject: Seminar - Model-Based Vision
[Forwarded from the MIT bboard by SASW@MIT-MC.]
W. ERIC L. GRIMSON
Local Constraints in
Model Based Recognition and Localization
From Sparse Data
April 23, 1984 4:00PM
NE43-8th floor playroom
A central characteristic of advanced applications in robotics is the
presence of significant uncertainty about the identities and attitudes
of objects in the workspace of a robot. The recognition and
localization of an object, from among a set of models, using sparse,
noisy sensory data can be cast as the search for a consistent matching
of the data elements to model elements. To minimize the computation,
local constraints are needed to limit the portions of the search space
that must be explicitly explored.
We derive a set of local geometric constraints for both the three
degree of freedom problem of isolated objects in stable positions, and
the general six degree of freedom problem of an object arbitrarily
oriented in space. We establish that the constraints are complete for
the case of three degrees of freedom, but not for six. We then show
by combinatorial analysis that the constraints are generally very
effective in restricting the search space and provide estimates for
the number of sparse data points needed to uniquely identify and
isolate the object. These results are supported by simulations of the
recognition technique under a variety of conditions that also
demonstrate its graceful degradation in the presence of noise. We
also discuss examples of the technique applied to real data from
several sensory modalities including laser ranging, sonar, and grey
level imaging.
Refreshments: 3:45PM
Host: Professor Patrick H. Winston
------------------------------
Date: Wed 18 Apr 84 14:43:58-PST
From: PENTLAND@SRI-AI.ARPA
Subject: Seminar - Robot Design Issues
[Forwarded from the Stanford bboard by Laws@SRI-AI.]
WHAT: FOUNDATIONAL ISSUES IN ROBOT DESIGN AND THEIR METHODOLOGICAL CONSEQUENCES
WHO: Stan Rosenschein, Artificial Intelligence Center, SRI International
SERIES: Issues in Language, Perception and Cognition
WHERE: Room 100, Psychology Dept.
WHEN: Monday April 23, 1:00pm <- * Note Change *
The design of software which would allow robots to exhibit complex
behavior in realistic physical environments is a central goal of
Artificial Intelligence (AI). In structuring its approaches to this
problem, AI has over the years been guided by a melange of concepts
from logic, computer programming, and (prominently) by certain
pretheoretic intuitions about mental life and its relationship to
physical events embodied in ordinary "folk psychology." This talk
presents two contrasting views of how information, perception, and
action might be modeled by a robot designer depending on how seriously
he took "folk psychology." One view takes the ascription of mental
properties to machines quite seriously and leads to a methodology in
which the abstract entities of folk psychology ("beliefs," "desires,"
"plans," "intentions", etc.) are realized in a one-for-one fashion as
data structures in the robot program. Frequently these data structures
resemble, in certain ways, the sentences of an interpreted logical
languages in that they are taken to express the "content" of the
belief, desire, etc. The alternative view does not assume this degree
of mental structure a priori. Logic may figure prominently, but it is
used chiefly BY THE DESIGNER to define and reason about the
environment and its relation to desired robot behavior. The talk will
suggest an automata-theoretic approach to the content of information
states which sidesteps many of the presuppositions of the folk
psychology. The implications of such an approach for a systematic
robot software methodology will be discussed, including the
possibility of "organism compilers." The thesis that AI's reliance on
folk psychology is, on balance, useful will be left unresolved though
certainly not unquestioned.
------------------------------
End of AIList Digest
********************
