Copy Link
Add to Bookmark
Report
Alife Digest Number 000
Welcome to the ARTIFICIAL LIFE Email Distribution List
------------------------------------------------------
At the second workshop for Artificial Life in Santa Fe, February 1990,
it was decided there was a need for an email distribution list to provide
a forum for discussion, ideas, comments and announcements in the area of
Artificial Life Research.
Artificial Life is the study of man-made systems that exhibit
behaviors characteristic of natural living systems. It
complements the traditional biological sciences concerned with
the analysis of living organisms by attempting to synthesize
life-like behaviors within computers and other
artificial media. By extending the empirical foundation
upon which biology is based beyond the carbon-chain life
that has evolved on Earth, Artificial Life can contribute
to theoretical biology by locating life-as-we-know-it
within the larger picture of life-as-it-could-be.
C. Langton 1989
The Artificial Life Research Group at Indiana University has been granted
the responsibility for maintaining and archiving this list.
Send all requests for additions/deletions/general business to:
alife-request@iuvax.cs.indiana.edu
Send all contributions to the alife list to:
alife@iuvax.cs.indiana.edu
Archives will soon be made available through anonymous FTP.
Sincerely,
Elisabeth Freeman
Eric Freeman
Marek Lugowski
Artificial Life Research Group
Computer Science Department
Indiana University, Bloomington
Article 7 of cs.alife:
Path: iuvax!marek
From: marek@iuvax.cs.indiana.edu (Marek Lugowski)
Newsgroups: cs.alife
Subject: impressions from Alife2?...
Message-ID: <35778@iuvax.cs.indiana.edu>
Date: 14 Feb 90 06:22:56 GMT
Sender: marek@iuvax.cs.indiana.edu
Lines: 12
Now might be a good time to begin asking for impressions of the Alife2
conference from those who are already hooked up to alife@iuvax, since
the list is still tiny and conducive to chats. I'd rather wait with
larger issues, such as of the edge of chaos, until we receive the
e-mail addresses Andi Sutherland is to give us shortly -- the bulk of
the initial membership. On the other hand, we need to test the list...
What have you taken home with you in a way of impressions, the 25 or
so of you out there?
-- Marek
marek@iuvax.cs.indiana.edu
Article 8 of cs.alife:
Path: iuvax!daemon
From: hiebeler@turing.cs.rpi.edu (Dave Hiebeler)
Newsgroups: cs.alife
Subject: Re: impressions from Alife2?...
Message-ID: <35816@iuvax.cs.indiana.edu>
Date: 14 Feb 90 16:57:40 GMT
Sender: daemon@iuvax.cs.indiana.edu
Lines: 14
I think it's going to take me several weeks to begin to sort out
the different ideas and perspectives that were presented at the
conference -- there was quite a lot of information to absorb.
Before the conference, I did promise the folks on Usenet that I would
try to post a summary of the conference, so I would like to hear a
bit of discussion about it here, to simplify that task.
So, should the proceedings be called "Life at the Edge of Chaos"? :-)
That certainly seems to be one of the stronger memes of the conference.
--
Dave Hiebeler hiebeler@turing.cs.rpi.edu
Computer Science Dept, RPI
Center for Nonlinear Studies, Los Alamos National Lab
Article 9 of cs.alife:
Path: iuvax!daemon
From: <cliddell@NMSU.Edu>
Newsgroups: cs.alife
Subject: alife2 conference
Message-ID: <35873@iuvax.cs.indiana.edu>
Date: 14 Feb 90 23:49:36 GMT
Sender: daemon@iuvax.cs.indiana.edu
Lines: 22
I am a biologist with a strong interest in many of the topics covered
at the alife2 conference. I was unable to attend the conference personally
but did manage to send a person from my lab. Many of the alife topics look
exciting to a real world biologist, for both theoretical and practical
applications. I am currently using L-systems and IFS to model plant growth
-not so unique- but my angle on this is to study the branching patterns
of fungi in order to aid in the identification of non-fruiting isolates.
This could have enormous practical benefits in the fields of microbiology,
mycology and plant pathology. My second interest is in applying these
methods to the dynamic aspects of biology, growth, metabolism etc as
opposed to descriptive biology. This prospect is exciting, although
certainly long term.
If anyone has an interest in the types of things I'm talking about. Get
in touch. I hope to be able to open some lines of communication between
biologists and the cs community that are not yet open.
Thanks
Craig Liddell
New Mexico State University
cliddell@nmsu.edu
Article 10 of cs.alife:
Path: iuvax!daemon
From: Stephen Smoliar <smoliar@vaxa.isi.edu>
Newsgroups: cs.alife
Subject: my Santa Fe trip report (submitted to my project leader)
Message-ID: <35919@iuvax.cs.indiana.edu>
Date: 15 Feb 90 14:47:41 GMT
Sender: daemon@iuvax.cs.indiana.edu
Lines: 103
Given that the Artificial Life Workshop held in Santa Fe on February 5-9 was
somewhat off the beaten trail of artificial intelligence, I think my spending
the week there deserves a trip report. However, before I try to address just
what this workshop was about, I would like to "set the stage" with a brief
account of the opening remarks by David Campbell, director of the Center for
Non-Linear Studies at Los Alamos. His basic theme that there were three
elements which were necessary for innovative science:
1. Diversity of ideas: This was basically inspired by his recently
having read Stephen Jay Gould's WONDERFUL LIFE, with its
description of the great diversity of life forms in the
Burgess shale and that argument that this diversity was
essential to the development of life on this planet.
2. Craziness: This one was inspired by Wolfgang Pauli and
observations that many elements of quantum theory were,
when first introduced, perceived as too crazy to be part
of any scientific theory.
3. Enthusiastic support: In other words, you can only get something
out of a broad diversity of ideas colored with craziness if
there is enough enthusiastic support to allow that crazy
diversity to flourish.
It was Campbell's hope that all three of these elements would prevail at this
workshop, and I think that hope was fulfilled. Certainly, they were more in
evidence here than at most traditional AI events.
So what, then, is "artificial life" all about? This was best discussed in the
overview of the conference which Christopher Langton offered after Campbell's
introduction. As Langton put it, the study of artificial life is based on the
hypothesis that what we call "life" is more a property of ORGANIZATION of
matter than of matter itself. Therefore, it would make sense to try to
implement that organization using some different form of matter. Such
an implementation would constitute artificial life by virtue of achieving
the same DYNAMICS OF INFORMATION that one finds in the organizational
properties of "real" life.
Why should this pursuit be of scientific interest? That answers are not that
different from the goals of artificial intelligence:
1. We want to build models through which we may better understand
"real" life.
2. We want to experiment with developing new systems.
As the workshop progressed, it emerged that there were two key elements which
artificial life systems strove to realize: adaptation and self-reproduction.
The study of adaptation could probably be viewed as a new approach to learning
and discovery (although since most connectionist work is based on adaptation,
the word "new" may be inappropriate). Richard Belew's discussion of this
aspect was particularly interesting for his attempt to demonstrate a close
coupling between learning and evolution. Also of interest was Danny Hillis'
demonstration of practical applications of adaptation which exceeded the
capabilities of most connectionist work. What made Danny's work interesting
was its focus on COEVOLUTION. What this means is that one should not focus
adaptation solely on the development of a better problem solver, which would
amount to an optimization problem over a suitably parameterized space.
Rather, that problem solver must adapt to the presence of a "parasite"
which is also evolving with the goal of attacking weak points in the
"fitness landscape" in which the problem solver is trying to find an
optimum point. Roughly speaking, as the problem solver tries to climb
a hill, the parasite tries to wear that hill down, forcing the problem
solver to search elsewhere. Such a system is not concerned with static
equilibrium but, rather, goes through periods of punctuated equilibrium
of the sort which Gould has envisaged for evolution. To demonstrate this
process, Hillis presented the discovery of a sort algorithm whose structure
was "attacked" by a parasite which tried to design test cases for which that
algorithm would fail.
Self-reproduction may not be viewed as immediately applicable to artificial
intelligence; but it is a property which computer science, as a whole, must
confront, since it is the primary property of "infections" like computer
viruses and worms. If we are going to deal effectively with such problems,
we need a better understanding of the self-reproductive properties.
Furthermore, we must be open to the possibility that the best remedy
for such problems may lie in the design of OTHER self-reproducing code
(dwelling on the metaphor of antibodies).
What both these elements have in common, and what Langton wished to stress,
is a focus on SYSTEM DYNAMICS. Pursuing the philosophy of Minsky's SOCIETY
OF MIND, there was a general recognition that knowledge is an emergent
property of system behavior. Indeed, Langton emphasized that artificial
intelligence was, in its early days, concerned with issues of biological
systems and that artificial life studies might benefit attempts for artificial
intelligence to return to some of its biological "roots." If we are to follow
this focus on system dynamics, then we should concern ourselves with attempts
to define the repertoire of actions which a given system may take. This
involves dealing with two primary questions:
1. What conditions allow for each of these actions?
2. What are the effects of each action?
As is the case with any workshop, there was variation of quality in the
presented material. However, I would argue that the DEGREE OF VARIATION
was much greater than we tend to find at artificial intelligence meetings.
Thus, there were broad intellectual differences between the very good and
the very bad, rather than minor deviations from an acceptable norm. While
this meant that the very bad could be quickly dismissed, it also meant that
the very good was also very thought-provoking; and it was quite exciting to
attend a meeting in which more than one or two events left me with many
questions which could constitute interesting research topics.
Article 11 of cs.alife:
Path: iuvax!daemon
From: Eric T. Freeman <efreeman@silver.ucs.indiana.edu>
Newsgroups: cs.alife
Subject: response to alife reference requests
Message-ID: <36021@iuvax.cs.indiana.edu>
Date: 16 Feb 90 04:44:08 GMT
Sender: daemon@iuvax.cs.indiana.edu
Lines: 26
Several people have asked for introductory references for Alife.
The Alife I proceedings are a good place to start:
Artificial Life: Proceedings of an Interdisciplinary Workshop
on the Synthesis and Simulation of Living Systems.
Edited by Christopher Langton
Volume VI
Hardback #09346-4
Paperback #09356-1
800-447-2226 (phone for addison-wesley pub.)
This volume contains two annotated bibliographies.
Eric Freeman
-----------------------------------------------------------------------------
Eric T. Freeman efreeman@silver.bacs.indiana.edu
Artificial Life Research Group freeman@dftnic.gsfc.nasa.gov
Indiana University Computer Science efreeman@cmns-sun.think.com
"Computer Science is not about computers, any more than astronomy is about
telescopes" - unknown
-----------------------------------------------------------------------------
Article 12 of cs.alife:
Path: iuvax!daemon
From: Ken Phillips <kp@media-lab.media.mit.edu>
Newsgroups: cs.alife
Subject: Artificial Life Workshop List
Message-ID: <36057@iuvax.cs.indiana.edu>
Date: 16 Feb 90 16:14:42 GMT
Sender: daemon@iuvax.cs.indiana.edu
Lines: 16
I would appreciate beingadded to the distribution / discussion
list:
kp@amt.mit.edu
(US Mail: Ken Phillips
41 Fifth Ave.
Suite 2E
New York, New York 10003
(212-477-4370)
thanks,
ken
Article 13 of cs.alife:
Path: iuvax!daemon
From: Bill Yakowenko <wjy@cs.arizona.edu>
Newsgroups: cs.alife
Subject: mailing list
Message-ID: <36111@iuvax.cs.indiana.edu>
Date: 16 Feb 90 21:25:15 GMT
Sender: daemon@iuvax.cs.indiana.edu
Lines: 6
Please put me on your AL mailing list. Thanks.
----====----====----====----====----====----====----====----====----====----
Bill Yakowenko "Too much reality can kill you."
(internet): wjy@cs.arizona.edu (uucp): {cmcl2,noao,allegra}!arizona!wjy
grad student, cs dept, u of a, tucson, az, usa, na, earth, sol, milky way
Article 14 of cs.alife:
Path: iuvax!daemon
From: L L Campbell <campbell@sun.acs.udel.edu>
Newsgroups: cs.alife
Subject: ALR List
Message-ID: <36112@iuvax.cs.indiana.edu>
Date: 16 Feb 90 21:38:53 GMT
Sender: daemon@iuvax.cs.indiana.edu
Lines: 14
Please add me to the mailing list for the Artificial Life Research Group.
Thank you.
L. Leon Campbell
University of Delaware
400 Morris
Newark, DE 19717-5267
(302) 451-6767
campbell@sun.acs.udel.edu
[Mail]&q
Article 15 of cs.alife:
Path: iuvax!daemon
From: <DJB85%ALBNYVMS.BITNET@UACSC2.ALBANY.EDU>
Newsgroups: cs.alife
Subject: Subscribe Don Byrd
Message-ID: <36118@iuvax.cs.indiana.edu>
Date: 16 Feb 90 22:18:27 GMT
Sender: daemon@iuvax.cs.indiana.edu
Lines: 2
Please add my name to the Alife list. Thanks. Don Byrd, State University
of New York at Albany.
Article 16 of cs.alife:
Path: iuvax!daemon
From: Elisabeth Freeman <bfreeman@silver.ucs.indiana.edu>
Newsgroups: cs.alife
Subject: Recent messages reposted
Message-ID: <36129@iuvax.cs.indiana.edu>
Date: 16 Feb 90 23:12:26 GMT
Sender: daemon@iuvax.cs.indiana.edu
Lines: 284
We have received so many requests to add people since the mailing list
started that I thought I would resend some of the initial messages.
If you have already read these, please ignore!
Beth Freeman
Artificial Life Research Group
Indiana University
-----------------------------------------------------------------------
Welcome to the ARTIFICIAL LIFE Email Distribution List
------------------------------------------------------
At the second workshop for Artificial Life in Santa Fe, February 1990,
it was decided there was a need for an email distribution list to provide
a forum for discussion, ideas, comments and announcements in the area of
Artificial Life Research.
Artificial Life is the study of man-made systems that exhibit
behaviors characteristic of natural living systems. It
complements the traditional biological sciences concerned with
the analysis of living organisms by attempting to synthesize
life-like behaviors within computers and other
artificial media. By extending the empirical foundation
upon which biology is based beyond the carbon-chain life
that has evolved on Earth, Artificial Life can contribute
to theoretical biology by locating life-as-we-know-it
within the larger picture of life-as-it-could-be.
C. Langton 1989
The Artificial Life Research Group at Indiana University has been granted
the responsibility for maintaining and archiving this list.
Send all requests for additions/deletions/general business to:
alife-request@iuvax.cs.indiana.edu
Send all contributions to the alife list to:
alife@iuvax.cs.indiana.edu
Archives will soon be made available through anonymous FTP.
Sincerely,
Elisabeth Freeman
Eric Freeman
Marek Lugowski
Artificial Life Research Group
Computer Science Department
Indiana University, Bloomington
Path: iuvax!marek
From: marek@iuvax.cs.indiana.edu (Marek Lugowski)
Subject: impressions from Alife2?...
Message-ID: <35778@iuvax.cs.indiana.edu>
Date: 14 Feb 90 06:22:56 GMT
Sender: marek@iuvax.cs.indiana.edu
Lines: 12
Now might be a good time to begin asking for impressions of the Alife2
conference from those who are already hooked up to alife@iuvax, since
the list is still tiny and conducive to chats. I'd rather wait with
larger issues, such as of the edge of chaos, until we receive the
e-mail addresses Andi Sutherland is to give us shortly -- the bulk of
the initial membership. On the other hand, we need to test the list...
What have you taken home with you in a way of impressions, the 25 or
so of you out there?
-- Marek
marek@iuvax.cs.indiana.edu
Path: iuvax!daemon
From: hiebeler@turing.cs.rpi.edu (Dave Hiebeler)
Subject: Re: impressions from Alife2?...
Message-ID: <35816@iuvax.cs.indiana.edu>
Date: 14 Feb 90 16:57:40 GMT
Sender: daemon@iuvax.cs.indiana.edu
Lines: 14
I think it's going to take me several weeks to begin to sort out
the different ideas and perspectives that were presented at the
conference -- there was quite a lot of information to absorb.
Before the conference, I did promise the folks on Usenet that I would
try to post a summary of the conference, so I would like to hear a
bit of discussion about it here, to simplify that task.
So, should the proceedings be called "Life at the Edge of Chaos"? :-)
That certainly seems to be one of the stronger memes of the conference.
--
Dave Hiebeler hiebeler@turing.cs.rpi.edu
Computer Science Dept, RPI
Center for Nonlinear Studies, Los Alamos National Lab
Path: iuvax!daemon
From: <cliddell@NMSU.Edu>
Subject: alife2 conference
Message-ID: <35873@iuvax.cs.indiana.edu>
Date: 14 Feb 90 23:49:36 GMT
Sender: daemon@iuvax.cs.indiana.edu
Lines: 22
I am a biologist with a strong interest in many of the topics covered
at the alife2 conference. I was unable to attend the conference personally
but did manage to send a person from my lab. Many of the alife topics look
exciting to a real world biologist, for both theoretical and practical
applications. I am currently using L-systems and IFS to model plant growth
-not so unique- but my angle on this is to study the branching patterns
of fungi in order to aid in the identification of non-fruiting isolates.
This could have enormous practical benefits in the fields of microbiology,
mycology and plant pathology. My second interest is in applying these
methods to the dynamic aspects of biology, growth, metabolism etc as
opposed to descriptive biology. This prospect is exciting, although
certainly long term.
If anyone has an interest in the types of things I'm talking about. Get
in touch. I hope to be able to open some lines of communication between
biologists and the cs community that are not yet open.
Thanks
Craig Liddell
New Mexico State University
cliddell@nmsu.edu
Path: iuvax!daemon
From: Stephen Smoliar <smoliar@vaxa.isi.edu>
Subject: my Santa Fe trip report (submitted to my project leader)
Message-ID: <35919@iuvax.cs.indiana.edu>
Date: 15 Feb 90 14:47:41 GMT
Sender: daemon@iuvax.cs.indiana.edu
Lines: 103
Given that the Artificial Life Workshop held in Santa Fe on February 5-9 was
somewhat off the beaten trail of artificial intelligence, I think my spending
the week there deserves a trip report. However, before I try to address just
what this workshop was about, I would like to "set the stage" with a brief
account of the opening remarks by David Campbell, director of the Center for
Non-Linear Studies at Los Alamos. His basic theme that there were three
elements which were necessary for innovative science:
1. Diversity of ideas: This was basically inspired by his recently
having read Stephen Jay Gould's WONDERFUL LIFE, with its
description of the great diversity of life forms in the
Burgess shale and that argument that this diversity was
essential to the development of life on this planet.
2. Craziness: This one was inspired by Wolfgang Pauli and
observations that many elements of quantum theory were,
when first introduced, perceived as too crazy to be part
of any scientific theory.
3. Enthusiastic support: In other words, you can only get something
out of a broad diversity of ideas colored with craziness if
there is enough enthusiastic support to allow that crazy
diversity to flourish.
It was Campbell's hope that all three of these elements would prevail at this
workshop, and I think that hope was fulfilled. Certainly, they were more in
evidence here than at most traditional AI events.
So what, then, is "artificial life" all about? This was best discussed in the
overview of the conference which Christopher Langton offered after Campbell's
introduction. As Langton put it, the study of artificial life is based on the
hypothesis that what we call "life" is more a property of ORGANIZATION of
matter than of matter itself. Therefore, it would make sense to try to
implement that organization using some different form of matter. Such
an implementation would constitute artificial life by virtue of achieving
the same DYNAMICS OF INFORMATION that one finds in the organizational
properties of "real" life.
Why should this pursuit be of scientific interest? That answers are not that
different from the goals of artificial intelligence:
1. We want to build models through which we may better understand
"real" life.
2. We want to experiment with developing new systems.
As the workshop progressed, it emerged that there were two key elements which
artificial life systems strove to realize: adaptation and self-reproduction.
The study of adaptation could probably be viewed as a new approach to learning
and discovery (although since most connectionist work is based on adaptation,
the word "new" may be inappropriate). Richard Belew's discussion of this
aspect was particularly interesting for his attempt to demonstrate a close
coupling between learning and evolution. Also of interest was Danny Hillis'
demonstration of practical applications of adaptation which exceeded the
capabilities of most connectionist work. What made Danny's work interesting
was its focus on COEVOLUTION. What this means is that one should not focus
adaptation solely on the development of a better problem solver, which would
amount to an optimization problem over a suitably parameterized space.
Rather, that problem solver must adapt to the presence of a "parasite"
which is also evolving with the goal of attacking weak points in the
"fitness landscape" in which the problem solver is trying to find an
optimum point. Roughly speaking, as the problem solver tries to climb
a hill, the parasite tries to wear that hill down, forcing the problem
solver to search elsewhere. Such a system is not concerned with static
equilibrium but, rather, goes through periods of punctuated equilibrium
of the sort which Gould has envisaged for evolution. To demonstrate this
process, Hillis presented the discovery of a sort algorithm whose structure
was "attacked" by a parasite which tried to design test cases for which that
algorithm would fail.
Self-reproduction may not be viewed as immediately applicable to artificial
intelligence; but it is a property which computer science, as a whole, must
confront, since it is the primary property of "infections" like computer
viruses and worms. If we are going to deal effectively with such problems,
we need a better understanding of the self-reproductive properties.
Furthermore, we must be open to the possibility that the best remedy
for such problems may lie in the design of OTHER self-reproducing code
(dwelling on the metaphor of antibodies).
What both these elements have in common, and what Langton wished to stress,
is a focus on SYSTEM DYNAMICS. Pursuing the philosophy of Minsky's SOCIETY
OF MIND, there was a general recognition that knowledge is an emergent
property of system behavior. Indeed, Langton emphasized that artificial
intelligence was, in its early days, concerned with issues of biological
systems and that artificial life studies might benefit attempts for artificial
intelligence to return to some of its biological "roots." If we are to follow
this focus on system dynamics, then we should concern ourselves with attempts
to define the repertoire of actions which a given system may take. This
involves dealing with two primary questions:
1. What conditions allow for each of these actions?
2. What are the effects of each action?
As is the case with any workshop, there was variation of quality in the
presented material. However, I would argue that the DEGREE OF VARIATION
was much greater than we tend to find at artificial intelligence meetings.
Thus, there were broad intellectual differences between the very good and
the very bad, rather than minor deviations from an acceptable norm. While
this meant that the very bad could be quickly dismissed, it also meant that
the very good was also very thought-provoking; and it was quite exciting to
attend a meeting in which more than one or two events left me with many
questions which could constitute interesting research topics.
Path: iuvax!daemon
From: Eric T. Freeman <efreeman@silver.ucs.indiana.edu>
Subject: response to alife reference requests
Message-ID: <36021@iuvax.cs.indiana.edu>
Date: 16 Feb 90 04:44:08 GMT
Sender: daemon@iuvax.cs.indiana.edu
Lines: 26
Several people have asked for introductory references for Alife.
The Alife I proceedings are a good place to start:
Artificial Life: Proceedings of an Interdisciplinary Workshop
on the Synthesis and Simulation of Living Systems.
Edited by Christopher Langton
Volume VI
Hardback #09346-4
Paperback #09356-1
800-447-2226 (phone for addison-wesley pub.)
This volume contains two annotated bibliographies.
Eric Freeman
-----------------------------------------------------------------------------
Eric T. Freeman efreeman@silver.bacs.indiana.edu
Artificial Life Research Group freeman@dftnic.gsfc.nasa.gov
Indiana University Computer Science efreeman@cmns-sun.think.com
"Computer Science is not about computers, any more than astronomy is about
telescopes" - unknown
-----------------------------------------------------------------------------
Article 17 of cs.alife:
Path: iuvax!daemon
From: acad!megalon!rudy@uunet.UU.NET (Rudy Rucker)
Newsgroups: cs.alife
Subject: ants
Message-ID: <36176@iuvax.cs.indiana.edu>
Date: 17 Feb 90 21:36:02 GMT
Sender: daemon@iuvax.cs.indiana.edu
Lines: 8
How many different ants programs are there now? It seems like so many peoplehave written them. I want to do an ants program myself soon, and I'd
like to have a clear idea of what's been done so far. If you have written
or seen an electronic antfarm, drop me a line.
rudy@autodesk.com
Rudy Rucker
Article 18 of cs.alife:
Path: iuvax!daemon
From: Cliff Joslyn <cjoslyn@bingvaxu.cc.binghamton.edu>
Newsgroups: cs.alife
Subject: Compatriot mailing list: CYBSYS-L
Message-ID: <36182@iuvax.cs.indiana.edu>
Date: 17 Feb 90 22:14:04 GMT
Sender: daemon@iuvax.cs.indiana.edu
Lines: 69
To my mind, the twin meta-disciplinary fields of Systems Science and
Cybernetics are at the base of the intellectual heritage of both AI and
ALife. Fortunately, Systems Science has been represented at the ALife
conferences from members of my department. I wish there was a more
conscious recognition of this heritage in the ALife movement, but this
is not a posting to whine. Rather, I'd like to inform this group of the
existence of a lively discussion group on Systems Science and
Cybernetics. I've advertised this list there, and we'd be glad to have
y'alls input over there.
O------------------------------------------------------------------------->
| Cliff Joslyn, Cybernetician at Large, cjoslyn@bingvaxu.cc.binghamton.edu
| Systems Science, SUNY Binghamton, Box 1070, Binghamton NY 13901, USA
V All the world is biscuit shaped. . .
============================================================================
ANNOUNCING FORMATION OF A MAILING LIST FOR SYSTEMS AND CYBERNETICS
An electronic mailing list dedicated to Systems Science and Cybernetics
is currently in operation on the SUNY-Binghamton computer system. The
list is commited to discussing a general understanding of the evolution
of complex, multi-level systems like organisms, minds, and societies as
informational entities containing possibly circular processes. Specific
subjects include Complex Systems Theory, Self-Organizing Systems Theory,
Dynamic Systems Theory, Artificial Intelligence, Network Theory,
Semiotics, fractal geometry, Fuzzy Set Theory, Recursive Theory, computer
simulation, Information Theory, and more.
The purposes of the list include: 1) facilitating discussion among those
working in or just interested in the general fields of Systems and
Cybernetics; 2) providing a means of communicating to the general
research community about the work that Systems Scientists and
Cyberneticians do; 3) housing a repository of electronic files for
general distribution concerning Systems and Cybernetics; and 4) providing
a central, public directory of working Systems Scientists and
Cyberneticians. The mailing list can store or transmit notes and
messages, technical papers, references, calls for papers, computer
programs, and pictures and diagrams.
The list is coordinated by members of the Systems Science department of
the Watson School at SUNY-Binghamton, and is affiliated with the
International Society for the Systems Sciences (ISSS) and the American
Society for Cybernetics (ASC). The list is open to everyone, and we
currently have two hundred members from America, Canada, and Europe. Our
subscribers are from both academia and industry, and while many are
active researchers, others are just "listening in". We share in an
exciting, ongoing, multi-way conversation about many aspects of Systems
and Cybernetics. Different levels and kinds of knowledge and experience
are represented.
We invite all to join the discussion. To subscribe, you need a computer
account with access to one of the international networks (e.g. BITNET,
USENET, ARPANET, INTERNET, CSNET). Send a file containing only the line:
'SUB CYBSYS-L Your Full Name' to the list server at the address
LISTSERV@BINGVMB.BITNET.
Once subscribed, please post a message to the list itself at the address
CYBSYS-L@BINGVMB.BITNET. In the message, include your name, affiliation,
and a brief description of your work and/or interest in the fields of
Systems and Cybernetics.
List moderator: CYBSYS@BINGVAXU.CC.BINGHAMTON.EDU
Author: Cliff Joslyn, CJOSLYN@BINGVAXU.CC.BINGHAMTON.EDU
Article 19 of cs.alife:
Path: iuvax!daemon
From: sane@cs.uiuc.edu (Aamod Sane)
Newsgroups: cs.alife
Subject: ants
Message-ID: <36204@iuvax.cs.indiana.edu>
Date: 18 Feb 90 06:22:53 GMT
Sender: daemon@iuvax.cs.indiana.edu
Lines: 6
What is an ants program? Is it a simulation of an antshill? Where can I get one
Thanks
Aamod Sane
sane@cs.uiuc.edu
Article 20 of cs.alife:
Path: iuvax!daemon
From: Aaron Sloman <aarons%cogs.sussex.ac.uk@NSFnet-Relay.AC.UK>
Newsgroups: cs.alife
Subject: AI, alife, and the space of possible designs
Message-ID: <36211@iuvax.cs.indiana.edu>
Date: 18 Feb 90 14:20:08 GMT
Sender: daemon@iuvax.cs.indiana.edu
Lines: 117
When I saw that the "alife" list had been set up I had some puzzles
about its scope. E.g. why doesn't it simply subsume ALL the existing
sub-groups concerned with AI, vision, language, neural nets, cognitive
science, etc.
So I asked to be added to the alife list for a trial period.
This led to a brief interchange with Marek Lugowski, who then suggested
I post some of my comments to the list. So here they are, extended a
little.
-----------------------------------------------------------------------
Marek had earlier said:
> well, be sure to help us by posting what you good people at Sussex are
> doing with Alife...
There's a sense in which ALL work in AI (of which there is quite a lot
at Sussex, of several different kinds, including work on natural
language, vision, intelligent tutoring systems, automatic program
analysis, planning, neural nets, genetic algorithms, motivation and
emotions, etc) appears to meet the definition given in your posting:
Artificial Life is the study of man-made systems that exhibit
behaviors characteristic of natural living systems. It
complements the traditional biological sciences concerned with
the analysis of living organisms by attempting to synthesize
life-like behaviors within computers and other
artificial media.
This is why I was uncertain of the scope of the list and wanted to
sample it.
As for this:
.....Artificial Life can contribute
to theoretical biology by locating life-as-we-know-it
within the larger picture of life-as-it-could-be.
I have for some time been talking and writing about the "space" of
possible designs, which contains both existing organisms and artefacts
and myriad other possible designs. E.g. in my paper in IJCAI85
entitled "What enables a machine to understand", I wrote:
|Instead of answering either 'YES' or 'NO' to the question whether
|suitably programmed computers can understand, we note that within the
|space of possible 'behaving systems' (including animals) there are
|infinitely many cases, some sharing more features with human minds, some
|fewer. The important task is to analyse the nature and the implications
|of these similarities and differences, without assuming existing English
|words can label the cases adequately.
[[I.e. we'll need new technical vocabulary to extend ordinary language]]
|
|....
|
|Instead of listing necessary and sufficient conditions for understanding
|I argued that there is a complex set of prototypical conditions,
|different subsets of which may be exemplified in different animals or
|machines, yielding a complex space of possible systems which we are only
|just beginning to explore. Our ordinary concepts, like 'understanding'
|are not suited to drawing global boundaries within such a space. At best
|we can analyse the implications of various different designs, and the
|capabilities they produce, or fail to produce.
Similar things can be said about 'intelligence', 'consciousness',
'emotions', etc.
I would also argue that the space of possible designs is not a
continuum, but has many discontinuities (at different levels), so that
one of our tasks is to identify the important design discontinuities and
examine the implications of crossing the boundaries. I believe this can
help to provide a new theoretical framework for comparative ethology,
studies of evolution, etc. But it's a huge task.
It won't be helped by philosophically motivated searches for _global_
dichotomies: e.g. asking "What divides those systems with consciousness
(or minds, or feelings, or understanding, etc) from those without?" This
diverts us from the more productive painstaking analysis of the myriad
_minor_ discontinuities (e.g. between different kinds of consciousness,
awareness, sentience, sensitivity, self-monitoring, understanding, or
whatever).
Implicit in this is a methodology for the study of the space of possible
designs for behaving systems, that eschews the search for "essences"
(e.g. what's the essence of consciousness, or understanding), and
promises to revolutionise folk-psychology by vastly extending its
vocabulary to cope with a host of cases that ordinary language was
never designed for.
We have hardly begun to grasp the variety of types of mechanisms
relevant to understanding all the imporant design possibilities for
behaving systems. It is quite likely that people looking back a couple of
hundred years from now will see that the totality of mechanisms studied
in the last quarter of this century both in AI and within the current
wave of enthusiasm for neural computation, addresses only a small subset
of what is needed for human-like intelligence.
Expanding and deepening our understanding will require BOTH bottom up
studies inspired by (e.g.) the variety of sub-structures and processes
to be found in human brains AND top down studies inspired by _detailed_
analysis of requirements for various sorts of human and animal
capabilities. Putting these bottom up and top down studies together
fruitfully, requires rare breadth of vision and scientific and technical
knowledge. It is not easy to train people for the task.
Fortunately the scientific community can compensate for limitations of
individual scientists, enabling even myopically motivated research to
be linked to broader issues when appropriate.
Is that what this list is about?
Aaron Sloman,
School of Cognitive and Computing Sciences,
Univ of Sussex, Brighton, BN1 9QH, England
EMAIL aarons@cogs.sussex.ac.uk
or:
aarons%uk.ac.sussex.cogs@nsfnet-relay.ac.uk
Article 21 of cs.alife:
Path: iuvax!daemon
From: Eric T. Freeman <efreeman@silver.ucs.indiana.edu>
Newsgroups: cs.alife
Subject: Welcome from ALRG and a few questions
Message-ID: <36217@iuvax.cs.indiana.edu>
Date: 18 Feb 90 17:06:42 GMT
Sender: daemon@iuvax.cs.indiana.edu
Lines: 34
Well in the last 3 days the size of the ALife mailing list has grown
from 3 to over 200. We are quite a crowd now.
We would like to take this time to urge everyone to post comments,
questions, material from their own personal interests which are
related to ALife. ALife research is extremely multi-disciplinary
and more fields are related than not.
I would also like to propose some questions over the next couple of
weeks for discussion. Many of these were brought up at the ALife conference.
Particularly in the panel discussions.
Here is the first question:
[1] How do we define Artificial Life and how will we know when (if) we
create it? Is there a turing-test for ALife? How close are
computer viruses to Alife? Since many computer viruses are malevolent
should we be more concerned with what we are creating and how we
create it? D. Farmer/C. Langton proposed that we should always develop
alife models is closed/controlled systems so that they can not `escape'.
Is this necessary? What the internet worm an bad example of not doing this?
ERic Freeman
-----------------------------------------------------------------------------
Eric T. Freeman efreeman@silver.bacs.indiana.edu
Artificial Life Research Group freeman@dftnic.gsfc.nasa.gov
Indiana University Computer Science efreeman@cmns-sun.think.com
"Computer Science is not about computers, any more than astronomy is about
telescopes" - unknown
-----------------------------------------------------------------------------
Article 22 of cs.alife:
Path: iuvax!daemon
From: Aamod Sane <sane@cs.uiuc.edu>
Newsgroups: cs.alife
Subject: Questions
Message-ID: <36218@iuvax.cs.indiana.edu>
Date: 18 Feb 90 17:54:27 GMT
Sender: daemon@iuvax.cs.indiana.edu
Lines: 39
[1] How do we define Artificial Life and how will we know when (if) we
create it? Is there a turing-test for ALife? How close are
Two characteristics all life, including borderline cases like TMV seems
to have is a drive/ability to survive, ability to (re)organise their environment
and response to stimuli.
computer viruses to Alife? Since many computer viruses are malevolent
What computer viruses lack is these things. They have been so far programmed to
"recognise" elements of a certain system only. Increasing the number of systems
they can recognise brings them no nearer. They do not, to my knowledge, identify
a system as being such.
should we be more concerned with what we are creating and how we
create it? D. Farmer/C. Langton proposed that we should always develop
alife models is closed/controlled systems so that they can not `escape'.
Is this necessary? What the internet worm an bad example of not doing this?
The onus for creating these things rests on the creators. I think this is
essential. One of the basic things essential for any species to survive is
a tendency to treat the unknown as dangerous. Intellingence grants you the
freedom to temper this by "Rational" analysis of the situation, so that this
tendency need not assume paranoid proportions, but a basic caution is
neccessary. That is the reason why we have, for instance, laws to restrict
the working of the pharmaceutical companies, new food products etc.
Another thing which would be nice is the development of some kind of
"standard"
anthropomorphism free vocabulary, as far as possible. Discussions about
these things have a tendency to degenerate into noise, especially when the
philosophizing (unrestrained) begins. The net worsens the inadequacy of
inter human communication, which is bad enough in re. these issues as it
is.
Aamod Sane
sane@cs.uiuc.edu
Article 23 of cs.alife:
Path: iuvax!daemon
From: Paul Pangaro <pan@athena.pangaro.dialnet.symbolics.com>
Newsgroups: cs.alife
Subject: ... a few questions
Message-ID: <36226@iuvax.cs.indiana.edu>
Date: 18 Feb 90 21:19:08 GMT
Sender: daemon@iuvax.cs.indiana.edu
Lines: 45
Date: 18 Feb 90 17:06:42 GMT
From: efreeman@silver.ucs.indiana.edu (Eric T. Freeman)
[...]
I would also like to propose some questions over the next couple of
weeks for discussion. Many of these were brought up at the ALife conference.
Particularly in the panel discussions.
Here is the first question:
[1] How do we define Artificial Life and how will we know when (if) we
create it? Is there a turing-test for ALife? How close are
computer viruses to Alife? Since many computer viruses are malevolent
should we be more concerned with what we are creating and how we
create it? D. Farmer/C. Langton proposed that we should always develop
alife models is closed/controlled systems so that they can not `escape'.
Is this necessary? What the internet worm an bad example of not doing this?
May I first state that although I believe I understand and have sympathy
for the stated goals of the "Artificial Life" stuff, I have even more
trouble accepting that phrase than I do "Artificial Intelligence." To
reject the qualifier "artificial" and merely say "either it is or it
isn't" seems glib but is, I believe, the essence of what we are about. I
would have thought the hassles over the term "AI" would have led others
to avoid it when coining a new one, but then, maybe those who coined
this one did not see the troubles of the other.
With that preamble I ask, rather than propose, whether Maturana, Varela
and Uribe's work on autopoiesis has been examined as the means to deal
with the issue of "What is life?" They specifically avoid (as, for
example, is explained quite straightforwardly in "Tree of Knowledge") a
list of attributes ("it reproduces, it takes in nutrients, it wiggles on
its belly like a reptile....") for "defining" life, and instead specify
that a living is one which produces its self [sic] by the processes that
make itself up (pardon the paraphrase of a rather more elaborate
concept).
I say I "ask" whether this work has been examined because I am on
slightly familiar with "Artificial Life" and do not wish to bring coals
to Newcastle. In any event, just as one must provide rebuttal to the
Turing Test if one disagrees with it, the concept of autopoiesis and
Maturana's "Biology of Cognition" deserves a response.
Best,
PANgaro
Article 24 of cs.alife:
Path: iuvax!silver!efreeman
From: efreeman@silver.ucs.indiana.edu (Eric T. Freeman)
Newsgroups: cs.alife
Subject: MAJOR CHANGE IN LOCAL ALIFE NEWSGROUP
Message-ID: <36228@iuvax.cs.indiana.edu>
Date: 18 Feb 90 22:57:15 GMT
Sender: root@iuvax.cs.indiana.edu
Reply-To: efreeman@silver.ucs.indiana.edu (Eric T. Freeman)
Distribution: cs
Organization: Indiana University, Bloomington
Lines: 12
Local postings TO cs.alife will no longer be broadcast to the
world-wide distribution list. To post to the distribution list
please mail to alife@iuvax. If you do post to cs.alife it will
only be seen locally.
Postings FROM the distribution list will still be forwarded to
cs.alife.
If you prefer to be on the distribution list rather than reading news
send mail to efreeman@silver (or bfreeman@silver or marek@iuvax).
ERic
Article 25 of cs.alife:
Path: iuvax!daemon
From: sztipaj@vuse.vanderbilt.edu (Janos Sztipanovits)
Newsgroups: cs.alife
Subject: Add me
Message-ID: <36231@iuvax.cs.indiana.edu>
Date: 19 Feb 90 00:37:17 GMT
Sender: daemon@iuvax.cs.indiana.edu
Lines: 6
Please add me to your mailing list.
Janos Sztipanovits
Vanderbilt University
Department of Electrical Engineering
Nashville, TN 37235
Article 26 of cs.alife:
Path: iuvax!daemon
From: <eiverson@NMSU.Edu>
Newsgroups: cs.alife
Subject: autocatalytic sets
Message-ID: <36235@iuvax.cs.indiana.edu>
Date: 19 Feb 90 00:59:02 GMT
Sender: daemon@iuvax.cs.indiana.edu
Lines: 60
I am assembling a bibliography on autocatalytic sets and string
matching/replication. Specifically, I am looking for ways in which
relatively primitive sets of strings of 4-6 elements in length can "bootstrap"
themselves in order to create longer, more complex strings. I see
this process as being similar to Markov chains; the main difference
being that the string assembly would be done in parallel and with
strings of increasingly longer lengths.
Currently, I am trying to apply this to monotonic music, where the
melody would be used as a string of elements. The string would then
be "broken down" by having "enzymes" bind to sites on the string and
break it in half. This would be done recursively until the original
string had been broken into small enough pieces. At this point, I
would like to reassemble the pieces until some sort of stable state is
reached. I would also like to introduce mutation and some sort of
fitness metric into this process; in order that the strings could
somehow "evolve." The fitness metric would here be defined as a
string's ability to catalyze other reactions. Strings unable to do
this would be excluded from the auto-catalytic set. Strings that were
able to catalyze reactions would be rewarded by being allowed to
replicate albeit with the possible hazzard of mutation.
Here are some references I've found so far:
Eigen, M. & Schuster, P. (1979) The Hypercycle, Springer-Verlag.
Eigen, M. (1971) "Self-Organization of Matter and Evolution of
Biological Macromolecules," Naturwissenshaften 58, 465.
Farmer, J. D. & Kauffman, S. A. & Packard, N. H. (1986) "Autocatalytic
Replication of Polymers," Physica 22D 50-67.
Farmer J. D. (????) "A Rosetta Stone for Connectionism," Los Alamos
National Laboratory.
Grassberger, P. (1986) "Toward a Quantitative Theory of Self-Generated
Complexity," Int. J. Theoret. Phys. 25, 907.
Kauffman, S. A. (1986) "Autocatalytic Sets of Proteins," J of Theor.
Biol. 119, 1-24.
Lindgren, K. (????) "Evolution in a Population of Mutating Strategies."
Molander, P. (1985) "The Optimal Level of Generosity in a Selfish,
Uncertain Environment," Journal of Conflict Resolution, Vol. 29 No. 4,
611-618.
Rasmussen S. & Knudsen, C. & Feldberg, R. & Hindsholm, M. (????) "The
Coreworld: Emergence and Evolution of Cooperative Structures in a
Computational Chemistry," Los Alamos National Laboratory.
Schuster, P. (1986) "Dynamics of Molecular Evolution," Physica 22D 100-119.
If anyone can think of additional references, or has a suggestion to
make, please write to:
Eric Iverson
eiverson@nmsu.edu
Article 27 of cs.alife:
Path: iuvax!daemon
From: acad!megalon!rudy@uunet.UU.NET (Rudy Rucker)
Newsgroups: cs.alife
Subject: ants
Message-ID: <36247@iuvax.cs.indiana.edu>
Date: 19 Feb 90 02:33:26 GMT
Sender: daemon@iuvax.cs.indiana.edu
Lines: 28
A couple of people asked what I mean by an "ants" program. Michael Travers
AGAR program is one, as is Jefferson's UCLA group's programs, as is
Dewdney's Turmite game, as is Chris Langton's Vants program. In all of these
you have a background of cells that may or may not ( cf. Norman Packard's
programs as well) act like a cellular automaton. Often the background
just sits there like paper that the ants can write on. The ants are
active cells. Unlike passive CA rules in which the cells update themselves
on basis of nearby neighbors, the ants are like ACTIVE CAs on top of
their background. The ants look at their neighborhood and then update
a neighbor cell by copying themselves into the neighbor. They copy
themselves either by moving their data into a neighbor, or by moving
a ponter to their (perhaps large) program into the neighbor.
What do ants do? In Turmites and Vants they simply sniff around and
turn this way and that according to trails they lay down. they act
like 2D turing machines. IN the next step up, they look for food and
either eat it or, more subtly, bring it back to a nest ( there was
a demo on this at alife II ). The next notch is that they reproduce
either sexually or asexually with or without mutation. The idea is
that, similar to Dewdney's Flibs, the ant programs can get better at
something --- finding food or following a trail.
One might also have ants which fight each other, cf. the "C Robots"
program of a couple of years ago where you put your own program in
with some others and they battle it out. It would be the most
interesting I think if the ants would build interesting structures.
Fractals, most likely.
Article 28 of cs.alife:
Path: iuvax!daemon
From: hiebeler@turing.cs.rpi.edu (Dave Hiebeler)
Newsgroups: cs.alife
Subject: ants
Message-ID: <36261@iuvax.cs.indiana.edu>
Date: 19 Feb 90 06:41:16 GMT
Sender: daemon@iuvax.cs.indiana.edu
Lines: 11
There is a group at Brussels that have done quite a bit of work
studying real ants and simulating their behavior. I have at least
one of their articles which was quite interesting and easy to read.
I'll enter the reference here when I get back to Troy in about a
week and a half, if no one else does so first. I don't have the
reference here with me in NM.
--
Dave Hiebeler hiebeler@turing.cs.rpi.edu
Computer Science Dept, RPI
Center for Nonlinear Studies, Los Alamos National Lab
Article 29 of cs.alife:
Path: iuvax!daemon
From: Aamod Sane <sane@cs.uiuc.edu>
Newsgroups: cs.alife
Subject: ants
Message-ID: <36262@iuvax.cs.indiana.edu>
Date: 19 Feb 90 07:19:04 GMT
Sender: daemon@iuvax.cs.indiana.edu
Lines: 25
There is a group at Brussels that have done quite a bit of work
studying real ants and simulating their behavior. I have at least
one of their articles which was quite interesting and easy to read.
I'll enter the reference here when I get back to Troy in about a
week and a half, if no one else does so first. I don't have the
reference here with me in NM.
--
Dave Hiebeler hiebeler@turing.cs.rpi.edu
Computer Science Dept, RPI
Center for Nonlinear Studies, Los Alamos National Lab
One thing interesting about the matter of ants is the ratio of "awareness"/
"intelligence" of the "hill" as a whole to that of an individual ant.
What if we consider an anthill with more "individualistic" ants,
i.e those which are not so physically functionalised that they cannot survive
as other than aggregates, but have a very strong bias toward living in a
community. Has one of the researching groups consider this.
Also of interest is human society/ individual ratios, where the intellingence
of society seems to be lower, that of individuals higher.
Aamod Sane
sane@cs.uiuc.edu
Article 30 of cs.alife:
Path: iuvax!daemon
From: V Bauchau/Dept Biologie, Louvain-la-Neuve <VINCENT%BUCLLN11.BITNET@CUNYVM.CUNY.EDU>
Newsgroups: cs.alife
Subject: Santa Fe Symposium
Message-ID: <36269@iuvax.cs.indiana.edu>
Date: 19 Feb 90 13:45:39 GMT
Sender: daemon@iuvax.cs.indiana.edu
Lines: 9
I would realy appreciate if somebody could post to the list a kind of
summary of the Artificial Life II congress held in Santa Fe.
Thank you very much,
//////////////// //////////////////////
Vincent Bauchau, Department of Biology,
Vincent@Buclln11 Univ. Louvain-la-Neuve
//////////////// //////////////////////
Article 31 of cs.alife:
Path: iuvax!daemon
From: fv@ccr-p6.univ-p6-7.fr (Francisco VARELA p3413)
Newsgroups: cs.alife
Subject: signup
Message-ID: <36327@iuvax.cs.indiana.edu>
Date: 19 Feb 90 18:28:59 GMT
Sender: daemon@iuvax.cs.indiana.edu
Lines: 6
Please sign me up for the A-life newsgroups. My address is:
fv@frunip62.BITNET.
Name: Francisco Varela, CNRS, Paris, France
Thanks,
--F.
Article 32 of cs.alife:
Path: iuvax!daemon
From: heirich%cs@ucsd.edu (Alan Heirich)
Newsgroups: cs.alife
Subject: Brussels/Self Organization/ALife
Message-ID: <36336@iuvax.cs.indiana.edu>
Date: 19 Feb 90 19:19:22 GMT
Sender: daemon@iuvax.cs.indiana.edu
Lines: 31
Dave Hiebeler (hiebeler@turing.cs.rpi.edu) writes:
There is a group at Brussels that have done quite a bit of work
studying real ants and simulating their behavior. I have at least
one of their articles which was quite interesting and easy to read.
I'll enter the reference here when I get back to Troy in about a
week and a half, if no one else does so first. I don't have the
reference here with me in NM.
I assume he is referring to the work of the "Brussels School", a collection
of biologists, molecular chemists, anthropologists, ... who study self
organization and collective behavior. They have a general interest in
self organization and the emergence of life-like behavior.
I would assume that people on this list will be interested in their work,
and in particular, in a NATO ASI workshop happening next month at the
University of Texas, Austin, organized by Ilya Prigogine, Roger Traub,
and several others. (I can psot more info if people are interested).
For those who are interested in their literature, the following references
may be helpful:
Jantsch, "The Self ORganizing Universe", 1976 (?)
Nicolis & Prigogine, "Self Organization in Non Equilibrium Systems", 1978
Weber, Depew & Smith, "Information, Entropy and Evolution", 1989
Alan Heirich
...................................................................
Comp. Sci. & Eng., Cognitive Sci. : heirich@cs.ucsd.edu
C-014, Univ. Calif. San Diego : aheirich@ucsd.BITNET
La Jolla, CA 92093 USA : (619) 546 6293
:::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::
Article 33 of cs.alife:
Path: iuvax!daemon
From: acad!megalon!rudy@uunet.UU.NET (Rudy Rucker)
Newsgroups: cs.alife
Subject: why ants?
Message-ID: <36370@iuvax.cs.indiana.edu>
Date: 19 Feb 90 22:11:46 GMT
Sender: daemon@iuvax.cs.indiana.edu
Lines: 30
Thanks for the interesting feedback on ants. I'm eager to get that
Brussels reference. A question which bothers me off and on is Why
Ants? That is, why does it seem so natural a step in creating
alife to work on ants. There are other steps of course, a whole
ladder of them, but something about ant colonies seems peculiarly
attractive in the field of alife. One reason is certainly that
ant colonies are made of lots of simple individuals and have
emergent behavior. I would suggest another reason is that ants are
a paradigm for an alife dream which I describe in my novels Software
and Wetware (Avon books paperback), the dream of a race of self-
replicating robots. The Software/Wetware robots, who are called
boppers, live on the Moon, and spend a lot of their time mining,
building factories, smelting chips, and assembling robot bodies
onto which they download their programs. It would be interesting
to try and consciously model this with an ants program. Have ants
that gather raw material and use it to build new ants. What is
the raw material they might gather? In a program, I think the
material would be the addresses of unused memory bytes. Think
of ants searching through your RAM finding extra memory regions that
can be used for copies of themselves. Sounds kind of virus-like,
actually.
One real risk to alife researchers is that some viral outrage will
so arouse public opinion that alife work becomes as heavily regulated
as gene-tinkering now is. Is there any way in which viral type
activities could be beneficial. One thought I once had was of
an ant operating system which makes copies of often used programs and
hides them for later use in case the original is damaged. Enter
the name of the program and get back the message "program is hiding.
is it safe to come out?" just a thought.
Article 34 of cs.alife:
Path: iuvax!daemon
From: Arshad Mahmood <arshad%lfcs.edinburgh.ac.uk@NSFnet-Relay.AC.UK>
Newsgroups: cs.alife
Subject: Re: AI, alife, and the space of possible designs
Message-ID: <36385@iuvax.cs.indiana.edu>
Date: 19 Feb 90 23:44:26 GMT
Sender: daemon@iuvax.cs.indiana.edu
Lines: 37
Prof. Sloman writes
>I would also argue that the space of possible designs is not a
>continuum, but has many discontinuities (at different levels), so that
>one of our tasks is to identify the important design discontinuities and
>examine the implications of crossing the boundaries. I believe this can
>help to provide a new theoretical framework for .....
I heared you give a seminar here in Edinburgh a month or so ago and am still
confused as to what you are referring to when you speak of
discontinuities and what is the meaning of the term 'different levels'.
Consider the situation with 'emotions', it would seem to me to be
improbable that a space representing such an emotion is discontinuous or
at least I cannot readily imagine it. It is perhaps possible for such
terms as 'understanding' since there is a great deal of evidence that lot's
of living things don't have this faculty, but since I don't know what I mean
by 'understanding' anyway the whole question seems a little ill formed.
I was however very intrigued by your enterprise to try and map small sections
of this space, how is this progressing ?
>Is that what this list is about ?
I suspect the list will become what people want of it. I personally would
like to see a very much experimental view of ALife discussed, as has already
been the case with the first set of postings and as was evident from the
proceedings of the first conference (I haven't seen the second). But then
again being an 'armchair' nanotech enthusiast I am hoping that someone will
come on and say they've just built an assembler.
Ash
Laboratory for the Foundations of Computer Science
Edinburgh University
Edinburgh
Scotland
Article 35 of cs.alife:
Path: iuvax!daemon
From: <alife-request@iuvax.cs.indiana.edu>
Newsgroups: cs.alife
Subject: AI, life and the space
Message-ID: <36399@iuvax.cs.indiana.edu>
Date: 20 Feb 90 01:25:21 GMT
Sender: daemon@iuvax.cs.indiana.edu
Lines: 43
Prof. Sloman writes
>I would also arguethat the space of possible designs is not a
>continuum, but has many discontinuities (at different levels), so that
>one of our tasks is to identify the important design discontinuities and
>examine the implications of crossing the boundaries. I believe this can
>help to provide a new theoretical framework for .....
I heared you give a seminar here in Edinburgh a month or so ago and am still
confused as to what you are referring to when you speak of
discontinuities and what is the meaning of the term 'different levels'.
What are the metrics that are used so as to define something like a design
space? There is a lot of similiarity in this situation and that of benchmarks
for computer architectures. Computer architectures are *extremely* simple
compared to the systems this group is discussing, yet witness the difficulty
in generating agreement on benchmarks or devising measures in that field.
again being an 'armchair' nanotech enthusiast I am hoping that someone will
come on and say they've just built an assembler.
Me too.
Ash
Laboratory for the Foundations of Computer Science
Edinburgh University
Edinburgh
Scotland
Aamod Sane
sane@cs.uiuc.edu
U. of Illinois at Urbana Champaign
--
--................................--................................--
.....Please bear with us while we "heal" alife's illin' software.
Your REPLY mechanism may now respond to: alife@iuvax.cs.indiana.edu
or to the individual sender, software willing. Please send carefully.
Article 36 of cs.alife:
Path: iuvax!daemon
From: <alife-request@iuvax.cs.indiana.edu>
Newsgroups: cs.alife
Subject: ALife: Question about 2nd Conf. etc.
Message-ID: <36401@iuvax.cs.indiana.edu>
Date: 20 Feb 90 01:32:35 GMT
Sender: daemon@iuvax.cs.indiana.edu
Lines: 23
I was wondering how the Second Conference (the one that just concluded at
Los Alamos) Proceedings were being distributed, for those of us unlucky enough
to have not been there :-(. Is this going to be another book in the Santa Fe
Institute Studies Series?
Secondly, I was going to be so bold as to suggest a convention where we, the
mailing group, could adopt beginning our Subject: line with the word Alife (as
done above) so as to indicate more generally the nature of the message.
-----------
Rajeev Pandey | Department of Computer Science
Internet: rpandey@cs.orst.edu | Computer Science Building 100
UUCP: hp-pcd!orstcs!rpandey | Oregon State University
Phone:(503) 737-3273 Fax: (503) 737-3014 | Corvallis, OR 97331-3902 U.S.A.
--
--................................--................................--
.....Please bear with us while we "heal" alife's illin' software.
Your REPLY mechanism may now respond to: alife@iuvax.cs.indiana.edu
or to the individual sender, software willing. Please send carefully.
Article 37 of cs.alife:
Path: iuvax!daemon
From: <alife-request@iuvax.cs.indiana.edu>
Newsgroups: cs.alife
Subject: why ants?
Message-ID: <36405@iuvax.cs.indiana.edu>
Date: 20 Feb 90 02:40:32 GMT
Sender: daemon@iuvax.cs.indiana.edu
Lines: 28
Why Ants? Maybe because since Sciences of the Artificial, not much has
has been accomplished with complex AI systems in simple environments :)
A nice thing to do with artificial ants would be to create
bidirectional communication channels between HOME and FOOD.
(Is this what was demo'ed at AILIFE II?)
Living with lots of natural ants a while ago, I remember how they
formed two-color (e.g. coming vs going) loops. Each would meander in a
pie-slice region forward from current position and direction, until
they met an ant (of opposite color), at which point they would reset
their position, direction, and deviation.
Initially short and swarmy, this would quickly organize into a very
straight and narrow material transport channel. The system was able to
rapidly recover from faults such as interruptions caused by large and
heavy leather-covered objects.
Jordan Pollack Assistant Professor
CIS Dept/OSU Laboratory for AI Research
2036 Neil Ave Email: pollack@cis.ohio-state.edu
Columbus, OH 43210 Fax/Phone: (614) 292-4890
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
WARNING: Your REPLY mechanism will now respond to either
alife@iuvax.cs.indiana.edu or to the individual sender, depending on the
implementation. PLEASE send carefully.
Article 38 of cs.alife:
Path: iuvax!daemon
From: <alife-request@iuvax.cs.indiana.edu>
Newsgroups: cs.alife
Subject: Nato ASI Workshop/Brussels School
Message-ID: <36413@iuvax.cs.indiana.edu>
Date: 20 Feb 90 03:51:38 GMT
Sender: daemon@iuvax.cs.indiana.edu
Lines: 95
Marek Lugowski (marek@iuvax.cs.indiana.edu) has asked me to post the
information about the Nato ASI workshop in March. Here is a notice
that I received some weeks ago. At first glance it may appear related
to machine learning more than to ALife, but in fact the research group
that has organized it is concerned with a wide variety of self-organizing
phenomena, including the emergence of life.
(I'm not affiliated with these people, just interested in their
work. For more info contact the address at the end of the message).
- Alan Heirich (heirich@cs.ucsd.edu)
===================================================================
Originally-From: R09614%BBRBFU01.BITNET@vma.CC.CMU.EDU
Subject: NATO Conference Announcement
Status: RO
ANNOUNCEMENT:
_______________________________________
NATO Advanced Research Workshop on
Self-organization, Emerging Properties
and Learning.
Center for Studies in Statistical Mechanics and
Complex Systems
The University of Texas
Austin, Texas, USA
March 12-14, 1990
_______________________________________
Topics
------
- Self-Organization and Dynamics in Networks of Interacting
Elements
- Dynamical Aspects of Neural Activity: Experiments and
Modelling
- From Statistical Physics to Neural Networks
- Role of Dynamical Attractors in Cognition and Memory
- Dynamics of Learning in Biological and Social Systems
The goal of the workshop is to review recent progress on self-
organization and the generation of spatio-temporal patterns in
multi-unit networks of interacting elements, with special
emphasis on the role of coupling and connectivity on the observed
behavior. The importance of these findings will be assessed from
the standpoint of information and cognitive sciences, and their
possible usefulness in the field of artificial intelligence will be
discussed. We will compare the collective behavior of model
networks with the dynamics inferred from the analysis of
cortical activity. This confrontation should lead to the design of
more realistic networks, sharing some of the basic properties of
real-world neurons.
Sponsors
--------
- NATO International Scientific Exchange Programmes
- International Solvay Institutes for Physics and Chemistry,
Brussels, Belgium
- Center for Statistical Mechanics and Complex Systems, The
University of Texas at Austin
- IC2 Institute of The University of Texas at Austin
International Organizing Committee
--------------------------------
Ilya Prigogine, The University of Texas at Austin and Free
University of Brussels
Gregoire Nicolis, Free University of Brussels
Agnes Babloyantz, Free University of Brussels
J. Demongeot, University of Grenoble, France
Linda Reichl, The University of Texas at Austin
Local Organizing Committee
-------------------------
Ilya Prigogine, George Kozmetsky, Ping Chen, Linda Reichl,
William Schieve, Robert Herman, Harry Swinney, Fred Phillips
For Further Information Contact:
-----------------------------
Professor Linda Reichl
Center for Statistical Mechanics
The University of Texas
Austin, TX 78712, USA
Phone: (512) 471-7253; Fax: (512) 471-9637;
Bitnet: CTAA450@UTA3081 or PAPE@UTAPHY
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
WARNING: Your REPLY mechanism will now respond to either
alife@iuvax.cs.indiana.edu or to the individual sender, depending on the
implementation. PLEASE send carefully.
Article 39 of cs.alife:
Path: iuvax!daemon
From: <alife-request@iuvax.cs.indiana.edu>
Newsgroups: cs.alife
Subject: why ants?
Message-ID: <36414@iuvax.cs.indiana.edu>
Date: 20 Feb 90 04:39:55 GMT
Sender: daemon@iuvax.cs.indiana.edu
Lines: 29
Date: Mon, 19 Feb 90 12:07:24 PST
From: acad!megalon!rudy@uunet.UU.NET (Rudy Rucker)
Thanks for the interesting feedback on ants. I'm eager to get that
Brussels reference. A question which bothers me off and on is Why
Ants?
Here are some of my reasons:
1) I'm interested in modelling cooperative behavior in general, and ants
were a good place to start.
2) I had some contacts at Harvard's MCZ Labs, where they do research on
real ants, and I was able to get some useful ideas and feedback.
3) In my modelling system, there is a loose but aesthetically pleasing
similarity between the group behavior of ants and the group behavior of
the behavioral modules (agents) within a single ant brain.
There's also the fact that our systems don't seem to be capable of
modelling anything more complicated than an insect to any degree of
detail. If you're going to spend time watching insects, the social
insects (ants, bees, and termites) are simply the most interesting.
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
WARNING: Your REPLY mechanism will now respond to either
alife@iuvax.cs.indiana.edu or to the individual sender, depending on the
implementation. PLEASE send carefully.
Article 40 of cs.alife:
Path: iuvax!daemon
From: <alife-request@iuvax.cs.indiana.edu>
Newsgroups: cs.alife
Subject: ref of paper on ants
Message-ID: <36426@iuvax.cs.indiana.edu>
Date: 20 Feb 90 09:20:44 GMT
Sender: daemon@iuvax.cs.indiana.edu
Lines: 33
This is a good review paper on ants from the "Brussels school":
Pasteels, J.M., Deneubourg, J.-L., Goss, S., 1987.
Transmission and amplification of information in a changing environment:
the case of the insect societies.
In: Laws of Nature and Human Conduct, Specificities and Unifying Themes.
Edited by Prigigine
Discoveries 1985 Symposium, Solvay, Bruxelles. Impr. Dieu-Brichart,
Ottignies-Louvain-la-Neuve, 129-155.
Pr JM Pasteels address is:
Universite Libre de Bruxelles
Laboratoire de Biologie Animale et Cellulaire
Avenue FD Roosevelt 50
B-1050 Bruxelles, Belgium
Note also a very recent book by Prigogine:
Nicolis & Prigogine
Exploring Complexity.
- Freeman
//////////////// //////////////////////
Vincent Bauchau, Department of Biology,
Vincent@Buclln11 Univ. Louvain-la-Neuve
//////////////// //////////////////////
--
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=---=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
Your REPLY mechanism may target: alife-request@iuvax.cs.indiana.edu,
alife@iuvax.cs.indiana.edu, the individual sender, or all three.
Please post carefully.
Article 41 of cs.alife:
Path: iuvax!daemon
From: <alife-request@iuvax.cs.indiana.edu>
Newsgroups: cs.alife
Subject: Why ants?
Message-ID: <36427@iuvax.cs.indiana.edu>
Date: 20 Feb 90 11:51:31 GMT
Sender: daemon@iuvax.cs.indiana.edu
Lines: 43
As a person who still hasn't gotten around to looking up autopoiesis
or autocatalytic (it is amazing the differing backgrounds one finds
in cross-disciplinary studies), I rejoice in the topic of ``Why Ants?''
I suspect that Simon's The Sciences of the Artificial had less to do
with it than Wilson's Insect Societies. For Simon, the interesting
thing about ants was that they went in truly bizzarre paths not
because of some internal complexity, but because of complexity of the
Earth's surface at their level of scale -- thus the complexity wasn't
in life, but in the non-living aspects of the world, which people
often think are less mysterious. Whereas, I believe that Artificial
Life enthusiasts are probably more interested in getting the ants to
interact than in producing complicated surfaces and trying to simulate
the ant's response to those surfaces.
Of course, Wilson is better known for bees than ants, but since bees
are relatively difficult to study because they fly around and we don't
(although I recall reading somewhere a year or so ago that someone in
Europe had built a bee/fly robot for such study), the more land-bound
ant is going to be easier to study (ignoring the centrality of the Fly,
ala Deither's To Know A Fly). Also, ants are officially interesting --
just look at any Ant Farm advertisement.
But perhaps the real motivation is the one that Gavin Miller reported
in his presentation of The Motion Dynamics of Snakes and Worms at
SIGGRAPH'88, i.e., he liked doing things like spiders and snakes because
although people are familiar with them, they don't really look that closely
at them.
It has been said that ants are as complicated as we can currently simulate.
But is that because we can really simulate ants or is that because our
understanding of ants has only reached a certain level of complication.
--- BOB (webber@athos.rutgers.edu ; rutgers!athos.rutgers.edu!webber)
--
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=---=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
Your REPLY mechanism may target: alife-request@iuvax.cs.indiana.edu,
alife@iuvax.cs.indiana.edu, the individual sender, or all three.
Please post carefully.
Article 42 of cs.alife:
Path: iuvax!daemon
From: <alife-request@iuvax.cs.indiana.edu>
Newsgroups: cs.alife
Subject: ants reference
Message-ID: <36431@iuvax.cs.indiana.edu>
Date: 20 Feb 90 14:20:49 GMT
Sender: daemon@iuvax.cs.indiana.edu
Lines: 28
In a brief flash of wisdom, I just remembered that last month, I started
putting together an on-line record of articles in my file. Which of course
means I could have looked up that reference all along. :-) Oh well,
here it is:
Authors: Frans Moyson, Bernard Manderick
Title: The Collective Behavior of Ants: An Example of Self-Organization
in Massive Parallelism
Source: AI Memo n 88-7, VUB AI Lab - Pleinlaan 2 - 1050 Brussels
Date: 1988
I don't remember where I actually obtained this thing.. I'm also
not 100% sure that's the article I was thinking of, but I'm pretty
sure (99%).
--
Dave Hiebeler hiebeler@turing.cs.rpi.edu
Computer Science Dept, RPI
Center for Nonlinear Studies, Los Alamos National Lab
--
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=---=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
Your REPLY mechanism may target: alife-request@iuvax.cs.indiana.edu,
alife@iuvax.cs.indiana.edu, the individual sender, or all three.
Please post carefully.
Article 43 of cs.alife:
Path: iuvax!daemon
From: <alife-request@iuvax.cs.indiana.edu>
Newsgroups: cs.alife
Subject: tech report available
Message-ID: <36454@iuvax.cs.indiana.edu>
Date: 20 Feb 90 16:43:50 GMT
Sender: daemon@iuvax.cs.indiana.edu
Lines: 50
*************** PLEASE DO NOT DISTRIBUTE TO OTHER LISTS ***************
The following technical report (CS-90-99) is available.
Requests for copies: library@cs.utk.edu
Other correspondence: maclennan@cs.utk.edu
-----------------------------------------------------------------------
Evolution of Communication in a Population of Simple Machines
Bruce MacLennan
Department of Computer Science
The University of Tennessee
Knoxville, TN 37996-1301
Technical Report CS-90-99
ABSTRACT
We show that communication may evolve in a population of simple
machines that are physically capable of sensing and modifying a
shared environment, and for which there is selective pressure in
favor of cooperative behavior. The emergence of communication
was detected by comparing simulations in which communication was
permitted with those in which it was suppressed. When communica-
tion was not suppressed we found that at the end of the experi-
ment the average fitness of the population was 84% higher and had
increased at a rate 30 times faster than when communication was
suppressed. Furthermore, when communication was suppressed, the
statistical association of symbols with situations was random, as
is expected. In contrast, permitting communication led to very
structured associations of symbols and situations, as determined
by a variety of measures (e.g., coefficient of variation and
entropy). Inspection of the structure of individual highly fit
machines confirmed the statistical structure. We also investi-
gated a simple kind of learning. This did not help when communi-
cation was suppressed, but when communication was permitted the
resulting fitness was 845% higher and increased at a rate 80
times as fast as when it was suppressed. We argue that the
experiments described here show a new way to investigate the
emergence of communication, its function in populations of simple
machines, and the structure of the resulting symbol systems.
--
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=---=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
Your REPLY mechanism may target: alife-request@iuvax.cs.indiana.edu,
alife@iuvax.cs.indiana.edu, the individual sender, or all three.
Please post carefully.
Article 44 of cs.alife:
Path: iuvax!daemon
From: <alife-request@iuvax.cs.indiana.edu>
Newsgroups: cs.alife
Subject: summary thoughts on conference
Message-ID: <36455@iuvax.cs.indiana.edu>
Date: 20 Feb 90 16:44:39 GMT
Sender: daemon@iuvax.cs.indiana.edu
Lines: 119
Marek Lugowski suggested that, because of the increase of the size of the
mailing list, it might be worth while for me to re-post the trip report I
filed on the Santa Fe Conference. Since I have seen a couple of requests
for summaries, I hope that this may fill the bill. I apologize to the early
subscribers who have seen this already.
*********************************************************************
Given that the Artificial Life Workshop held in Santa Fe on February 5-9 was
somewhat off the beaten trail of artificial intelligence, I think my spending
the week there deserves a trip report. However, before I try to address just
what this workshop was about, I would like to "set the stage" with a brief
account of the opening remarks by David Campbell, director of the Center for
Non-Linear Studies at Los Alamos. His basic theme that there were three
elements which were necessary for innovative science:
1. Diversity of ideas: This was basically inspired by his recently
having read Stephen Jay Gould's WONDERFUL LIFE, with its
description of the great diversity of life forms in the
Burgess shale and that argument that this diversity was
essential to the development of life on this planet.
2. Craziness: This one was inspired by Wolfgang Pauli and
observations that many elements of quantum theory were,
when first introduced, perceived as too crazy to be part
of any scientific theory.
3. Enthusiastic support: In other words, you can only get something
out of a broad diversity of ideas colored with craziness if
there is enough enthusiastic support to allow that crazy
diversity to flourish.
It was Campbell's hope that all three of these elements would prevail at this
workshop, and I think that hope was fulfilled. Certainly, they were more in
evidence here than at most traditional AI events.
So what, then, is "artificial life" all about? This was best discussed in the
overview of the conference which Christopher Langton offered after Campbell's
introduction. As Langton put it, the study of artificial life is based on the
hypothesis that what we call "life" is more a property of ORGANIZATION of
matter than of matter itself. Therefore, it would make sense to try to
implement that organization using some different form of matter. Such
an implementation would constitute artificial life by virtue of achieving
the same DYNAMICS OF INFORMATION that one finds in the organizational
properties of "real" life.
Why should this pursuit be of scientific interest? That answers are not that
different from the goals of artificial intelligence:
1. We want to build models through which we may better understand
"real" life.
2. We want to experiment with developing new systems.
As the workshop progressed, it emerged that there were two key elements which
artificial life systems strove to realize: adaptation and self-reproduction.
The study of adaptation could probably be viewed as a new approach to learning
and discovery (although since most connectionist work is based on adaptation,
the word "new" may be inappropriate). Richard Belew's discussion of this
aspect was particularly interesting for his attempt to demonstrate a close
coupling between learning and evolution. Also of interest was Danny Hillis'
demonstration of practical applications of adaptation which exceeded the
capabilities of most connectionist work. What made Danny's work interesting
was its focus on COEVOLUTION. What this means is that one should not focus
adaptation solely on the development of a better problem solver, which would
amount to an optimization problem over a suitably parameterized space.
Rather, that problem solver must adapt to the presence of a "parasite"
which is also evolving with the goal of attacking weak points in the
"fitness landscape" in which the problem solver is trying to find an
optimum point. Roughly speaking, as the problem solver tries to climb
a hill, the parasite tries to wear that hill down, forcing the problem
solver to search elsewhere. Such a system is not concerned with static
equilibrium but, rather, goes through periods of punctuated equilibrium
of the sort which Gould has envisaged for evolution. To demonstrate this
process, Hillis presented the discovery of a sort algorithm whose structure
was "attacked" by a parasite which tried to design test cases for which that
algorithm would fail.
Self-reproduction may not be viewed as immediately applicable to artificial
intelligence; but it is a property which computer science, as a whole, must
confront, since it is the primary property of "infections" like computer
viruses and worms. If we are going to deal effectively with such problems,
we need a better understanding of the self-reproductive properties.
Furthermore, we must be open to the possibility that the best remedy
for such problems may lie in the design of OTHER self-reproducing code
(dwelling on the metaphor of antibodies).
What both these elements have in common, and what Langton wished to stress,
is a focus on SYSTEM DYNAMICS. Pursuing the philosophy of Minsky's SOCIETY
OF MIND, there was a general recognition that knowledge is an emergent
property of system behavior. Indeed, Langton emphasized that artificial
intelligence was, in its early days, concerned with issues of biological
systems and that artificial life studies might benefit attempts for artificial
intelligence to return to some of its biological "roots." If we are to follow
this focus on system dynamics, then we should concern ourselves with attempts
to define the repertoire of actions which a given system may take. This
involves dealing with two primary questions:
1. What conditions allow for each of these actions?
2. What are the effects of each action?
As is the case with any workshop, there was variation of quality in the
presented material. However, I would argue that the DEGREE OF VARIATION
was much greater than we tend to find at artificial intelligence meetings.
Thus, there were broad intellectual differences between the very good and
the very bad, rather than minor deviations from an acceptable norm. While
this meant that the very bad could be quickly dismissed, it also meant that
the very good was also very thought-provoking; and it was quite exciting to
attend a meeting in which more than one or two events left me with many
questions which could constitute interesting research topics.
--
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=---=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
Your REPLY mechanism may target: alife-request@iuvax.cs.indiana.edu,
alife@iuvax.cs.indiana.edu, the individual sender, or all three.
Please post carefully.
Article 45 of cs.alife:
Path: iuvax!marek
From: The Artificial Life Mailing List <marek@iuvax.cs.indiana.edu>
Newsgroups: cs.alife
Subject: archives, coming up!
Message-ID: <36457@iuvax.cs.indiana.edu>
Date: 20 Feb 90 17:12:27 GMT
Sender: marek@iuvax.cs.indiana.edu
Lines: 17
yes, we are hugging the edge of chaos at alife@iuvax, and an ftp
facility with archives and a means of posting (you posting, that is)
whatever you desire -- anyway, we hope to have all of these things
emerge from the liquid phase, a tad, oh, by midnight pizza time?...
Bear with us...
-- Marek
Alife R G, IU
--
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=---=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
Your REPLY mechanism may target: alife-request@iuvax.cs.indiana.edu,
alife@iuvax.cs.indiana.edu, the individual sender, or all three.
Please post carefully.
Article 46 of cs.alife:
Path: iuvax!daemon
From: <alife-request@iuvax.cs.indiana.edu>
Newsgroups: cs.alife
Subject: Re: AI, alife, and the space of possible designs
Message-ID: <36462@iuvax.cs.indiana.edu>
Date: 20 Feb 90 17:36:21 GMT
Sender: daemon@iuvax.cs.indiana.edu
Lines: 88
As one who currently regards artificial intelligence as his "home base," I
would like to try to respond to the following comment by Aaron Sloman:
> When I saw that the "alife" list had been set up I had some puzzles
> about its scope. E.g. why doesn't it simply subsume ALL the existing
> sub-groups concerned with AI, vision, language, neural nets, cognitive
> science, etc.
Certainly, to the extent that there are life forms which exhibit various levels
of cognition, such subsumption would seem to be valid. However, I'm not sure
that the issue is one of trying to lay out a hierarchy of scientific pursuits.
Physics may subsume all of chemistry; but that does not obviate, say, organic
chemistry as a discipline in its own right.
The reason I was drawn to the conference in Santa Fe was because, now that it
appears that artificial intelligence is become more aware of valuable
connections to psychology, there seems to be a need for heightened awareness
of biology, as well. There are, of course, members of the community who have
had such an awareness. Indeed, to the extent that cybernetics may be viewed
as a forerunner of artificial intelligence, it should also be seen as a domain
of inquiry about the dynamics of organic systems and how such systems may be
modeled or developed.
In his closing remarks to the Santa Fe Conference, Chris Langton observed that
artificial intelligence made what could be called a major paradigm commitment
in its early days by focusing its effort on SYMBOLIC processes. The recent
rise of connectionism may be viewed as a challenge to this paradigm; but, for
the most part, the conflict is being waged on the arena of the implementation
of specific artifacts. There are very few neural net researchers who are
primarily interested in real neurons, what they do, and how they contribute
to the behavior of the overall organism.
However, there are several active areas of inquiry which would appear to be
intimately related to biology and may thus benefit from the pursuits of
artificial life. For my part, I am most interested in the work of Gerald
Edelman. In NEURAL DARWINISM he introduced arguments for mechanisms of
selection as a means by which an organisms would acquire the ability to
classify images. This was demonstrated with models which were simple
enough to implement as software but complex enough to demonstrate his
thesis. Thus, to draw upon a primary theme of the Artificial Life Conference
(as I summarized it in my trip report, which I am willing to promote
shamelessly), this was an approach to the ability to detect and classify
patterns through an ADAPTIVE mechanism. In this same book Edelman also
took on issues of SELF-REPLICATION, the other primary theme of the conference;
and this theme was developed further in his following book, TOPOBIOLOGY. Now
Edelman has a third book, THE REMEMBERED PRESENT: A BIOLOGICAL THEORY OF
CONSCIOUSNESS. He is now bold enough to move on to general questions of
memory and language; and I, for one, am curious to see where this will
lead. My hope is that it will lead to further insights as to how thinking
about biology may contribute to concrete tasks we had set out for ourselves
in artificial intelligence.
Another key source of biology-based thinking is the recent "society of mind"
work of Marvin Minsky. I am willing to confess that it was only after becoming
comfortable with Edelman's biological view of the world that I began to develop
a sense of how to read many of the things Minsky had to say in THE SOCIETY OF
MIND and the papers which led up to it. I was also encouraged to see that there
were several participants at the Santa Fe Conference who were considering
issues of implementation of some of Minsky's ideas. While Minsky's seminal
thoughts about frames could be picked up by any industrious hacker with a good
sense of data management, it would seem that implementing the sorts of agents
which now interest Minsky requires some biological intuition as well.
Thus, subsumption is not the issue; INFLUENCE is. Just as cognitive science
encouraged a new breed of AI hackers to start learning something about
psychology, artificial life may encourage them to look at biology, too.
Whether or not they will gain from this new perspective remains to be
seen; but I, for one, am glad to see the influence kick in.
I suppose this is also a good time to comment on the call from the cybernetics
camp from Cliff Joslyn. I am willing to confess openly that I have always
shied away from cybernetics publications because they have, in the past, struck
me as being TOO diffuse. This is why I hope we can get away from arguments
about subsumption among disciplines. I view artificial life as a forum in
which I can learn things about investigations concerned with "artificial"
implementations of biological phenomena in the hope that some of those
implementations may provide insight into my attempts to implement systems
with flexible powers of memory. I would worry if this forum began to get
muddled with all the different issues of systems theory which have been
addressed by cybernetics. After all, I have enough mail to read already!
--
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=---=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
Your REPLY mechanism may target: alife-request@iuvax.cs.indiana.edu,
alife@iuvax.cs.indiana.edu, the individual sender, or all three.
Please post carefully.
Article 47 of cs.alife:
Path: iuvax!daemon
From: <alife-request@iuvax.cs.indiana.edu>
Newsgroups: cs.alife
Subject: Re: why ants?
Message-ID: <36463@iuvax.cs.indiana.edu>
Date: 20 Feb 90 18:04:19 GMT
Sender: daemon@iuvax.cs.indiana.edu
Lines: 18
esp: what viruses are good for.
I think that some of the most probable uses for "good" viruses would be
memory (primary and secondary) scavenging. Another "good" use might be the
care and maintenance of networks. you could have a virus that reproduces only
when network connections fail. On failure, you'd get a "swarm" of viruses at
the faulty port. If sysops weren't attentive, it'd eventrually impact
performance, and GET sysops attention.
uxmal
--
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=---=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
Your REPLY mechanism may target: alife-request@iuvax.cs.indiana.edu,
alife@iuvax.cs.indiana.edu, the individual sender, or all three.
Please post carefully.
Article 48 of cs.alife:
Path: iuvax!daemon
From: <alife-request@iuvax.cs.indiana.edu>
Newsgroups: cs.alife
Subject: Re: ants
Message-ID: <36464@iuvax.cs.indiana.edu>
Date: 20 Feb 90 18:19:34 GMT
Sender: daemon@iuvax.cs.indiana.edu
Lines: 68
>From sane@cs.uiuc.edu Sun Feb 18 22:41:25 1990
>
>One thing interesting about the matter of ants is the ratio of "awareness"/
>"intelligence" of the "hill" as a whole to that of an individual ant.
I don't know of any biologist that has attributed awareness or intelligence
to ants, either as individuals or as colonies. I hope this is what you mean
by putting quotes around these ill-defined words.
>What if we consider an anthill with more "individualistic" ants,
>i.e those which are not so physically functionalised that they cannot survive
>as other than aggregates, but have a very strong bias toward living in a
>community.
In primitive ants, all ants in a colony are identical, although generally
only one individual actually reproduces. In more specialized species,
there are two distinct morphological castes: reproductives and workers.
The reproductives probably can survive on their own (that is how they
start a new colony) for a while and reproduce successfully, and the workers
can live without a queen (although they can only produce males--no females).
In the most highly specialize species, the queen cannot live long without
workers, and the workers truly are sterile. In addition, there are often
multiple morphologically different worker castes, each specialized to a
specific task.
There is no reason why an individual ant cannot just walk away from the
colony, and try to make it on it's own. In fact, this may happen on
occasion.
>Has one of the researching groups consider this.
At UCLA, we have not addressed this question. In my view, this is not an
appropriate question (at least in terms of ants). In general, I think
that the reason people study ants is to look at eusocial behavior and
organization. The eusocial system of ants is so successful that it is
almost certain that an individual ant will have a much lower fitness
outside of the colony.
Another argument against this idea (again in terms of ants) is the fact
that an ant colony is often best viewed as a single organism. With this
view, an individual ant striking out on it's own (and dying in the process)
is similar to you having some of your skin scratched away--it is a very
minor inconvienience. One can also view the colony with the queen being
the organism of interest, and the workers being automata that she has
produced to serve her interests.
>Also of interest is human society/ individual ratios, where the intellingence
>of society seems to be lower, that of individuals higher.
I think intelligence plays a very minor role in real life, and probably
should also play a minor role in ALife as well. Any emphasis we see
on intelligence is only an artifact of the number of AI people that see
ALife as another name for AI.
>
>Aamod Sane
>sane@cs.uiuc.edu
Rob Collins
rjc@cs.ucla.edu
--
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=---=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
Your REPLY mechanism may target: alife-request@iuvax.cs.indiana.edu,
alife@iuvax.cs.indiana.edu, the individual sender, or all three.
Please post carefully.
Article 49 of cs.alife:
Path: iuvax!daemon
From: <alife-request@iuvax.cs.indiana.edu>
Newsgroups: cs.alife
Subject: Re: Why ants?
Message-ID: <36469@iuvax.cs.indiana.edu>
Date: 20 Feb 90 18:53:04 GMT
Sender: daemon@iuvax.cs.indiana.edu
Lines: 40
>Of course, Wilson is better known for bees than ants, but since bees
>are relatively difficult to study because they fly around and we don't
>(although I recall reading somewhere a year or so ago that someone in
>Europe had built a bee/fly robot for such study), the more land-bound
>ant is going to be easier to study (ignoring the centrality of the Fly,
>ala Deither's To Know A Fly). Also, ants are officially interesting --
>just look at any Ant Farm advertisement.
Gee, I thought Wilson was better know for ants! Maybe that is just my
bias. BTW, keep your eyes open for his new book. Unlike Insect Societies,
it will be exclusively about ants (or so I have been told).
>It has been said that ants are as complicated as we can currently simulate.
>But is that because we can really simulate ants or is that because our
>understanding of ants has only reached a certain level of complication.
I think that anyone who is serious about ALife already knows that we have
a long way to go before we can simulate even a single bacteria with
reasonable accuracy. What we (the crazy ant simulator people) have
looked at is foraging behavior and simple chemical communication. To
my knowledge, nobody has tried to simulate anything inside the nest
(or the building of the nest!). Of course, that is where all the
interesting stuff happens--we just have little or no idea of what really
goes on in there. A simulation like what Michael Travers demoed certainly
has at least the qualitative feel of foraging/communicating/cooperating
ants. For my work, I am trying to examine how and why such complex
behavior might have evolved.
>--- BOB (webber@athos.rutgers.edu ; rutgers!athos.rutgers.edu!webber)
rob collins
rjc@cs.ucla.edu
--
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=---=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
Your REPLY mechanism may target: alife-request@iuvax.cs.indiana.edu,
alife@iuvax.cs.indiana.edu, the individual sender, or all three.
Please post carefully.
Article 50 of cs.alife:
Path: iuvax!daemon
From: <alife-request@iuvax.cs.indiana.edu>
Newsgroups: cs.alife
Subject: software
Message-ID: <36479@iuvax.cs.indiana.edu>
Date: 20 Feb 90 20:14:17 GMT
Sender: daemon@iuvax.cs.indiana.edu
Lines: 28
Regarding my talk at alife II...
I forgot to mention that I am perfectly
willing to give away the software to anyone
interested in carrying out experiments on the
model biosphere. The simulation basically
consists of:
A World
-- 2D
-- with Food (energy) field
and Bugs
--with genes
--mutable move algorithm
--Sex
Norman Packard
n@complex.ccsr.uiuc.edu
--
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=---=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
Your REPLY mechanism may target: alife-request@iuvax.cs.indiana.edu,
alife@iuvax.cs.indiana.edu, the individual sender, or all three.
Please post carefully.
Article 51 of cs.alife:
Path: iuvax!daemon
From: <alife-request@iuvax.cs.indiana.edu>
Newsgroups: cs.alife
Subject: archive, software & paper depository available
Message-ID: <36492@iuvax.cs.indiana.edu>
Date: 20 Feb 90 21:11:08 GMT
Sender: daemon@iuvax.cs.indiana.edu
Lines: 43
ANNOUNCING ALIFE PAPER, SOFTWARE, AND ARCHIVE DEPOSITORY
--------------------------------------------------------
Speaking of software, papers, and archives we are now ready to accept
all of these through anonymous FTP. This is set up on the same
machine as the alife list (iuvax).
Current directories are:
~ftp/pub/alife/papers papers, tech reports, etc.
~ftp/pub/alife/software alife programs, source etc.
~ftp/pub/alife/archives archives of the mailing list
~ftp/pub/alife/public place for YOU to deposit
incoming materials.
machine: iuvax.cs.indiana.edu
username: anonymous
password: guest
Authors please attach your own copyright-type messages if applicable.
Support for readers without access to the INTERNET is unclear at this
time. We really need more person-power before we can commit to
answering mail requests for these items. If you have any questions send to
alife-request@iuvax.cs.indiana.edu
ERic
-----------------------------------------------------------------------------
Eric T. Freeman efreeman@silver.bacs.indiana.edu
Artificial Life Research Group freeman@dftnic.gsfc.nasa.gov
Indiana University Computer Science efreeman@cmns-sun.think.com
"Computer Science is not about computers, any more than astronomy is about
telescopes" - unknown
-----------------------------------------------------------------------------
--
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=---=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
Your REPLY mechanism may target: alife-request@iuvax.cs.indiana.edu,
alife@iuvax.cs.indiana.edu, the individual sender, or all three.
Please post carefully.
Article 52 of cs.alife:
Path: iuvax!daemon
From: <alife-request@iuvax.cs.indiana.edu>
Newsgroups: cs.alife
Subject: (none)
Message-ID: <36493@iuvax.cs.indiana.edu>
Date: 20 Feb 90 21:11:50 GMT
Sender: daemon@iuvax.cs.indiana.edu
Lines: 31
Subject: TR Available - Connectionist Language Users
Connectionist Language Users
Robert B. Allen
Bell Communications Research
The Connectionist Language Users (CLUES) paradigm employs neural
learning algorithms to develop grounded reactive intelligent
agents. In much of the research reported here, these agents
"use" language to answer questions and to interact with their en-
vironment. The model is applied to simple examples of generating
verbal descriptions, answering questions, pronoun reference, la-
beling actions, and verbal interactions between agents. In addi-
tion the agents are shown to be able to model other intelligent
activities such as planning, grammars and simple analogies, and
an adaptive pedagogy is introduced. Overall, these networks pro-
vide a natural account of many aspects of language use.
______
Request reports from smk@bellcore.com, Selma Kaufman, 2M356, Bell
Communications Research, Morristown, NJ 07960-1910
--
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=---=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
Your REPLY mechanism may target: alife-request@iuvax.cs.indiana.edu,
alife@iuvax.cs.indiana.edu, the individual sender, or all three.
Please post carefully.
