Overview of Tierra at ATR

bankpottstownAI and Robotics

Oct 23, 2013 (3 years and 10 months ago)

94 views

Overview of Tierra at ATR


Origin of Tierra


Ray, T. S. 1991. An approach to the synthesis of life. In: Langton, C., C. Taylor, J. D. Farmer, & S.
Rasmussen [eds], Artificial Life II, Santa Fe Institute Studies in the Sciences of Complexity,
vol. XI, 371
-
4
08. Redwood City, CA: Addison
-
Wesley.

Tierra originated at the end of 1989, a few years before coming to ATR. The creation of
Tierra was motivated by a desire to observe the evolutionary process in a medium other than carbon
chemistry. This motivation i
s related to the desire to observe life on other planets. If we could
observe life on another planet, it would be another instance of evolution, completely independent
from the evolution of life on Earth. Learning about life on other planets would broade
n our
knowledge and understanding of both life and the evolutionary process that generates life. Our
current knowledge of life and evolution is based on a sample size of one: life on Earth.

The prospects for observing life on another planet in our lifetim
es are remote. Yet, we are
discovering that we can create genuine instances of some of life's processes in artificial media, such
as in the computer. These new instantiations of life processes are known as Artificial Life, and
Tierra is an example of ins
tantiating evolution in the digital medium. It is a particularly significant
example of Artificial Life because evolution is a defining property of life, and the creative process
that generates life.

Co
-
evolutionary dynamic
-

expanding fitness landscape

P
erhaps the most dramatic result of the original Tierra experiment was the active co
-
evolutionary dynamic. Previous artificial evolving systems were not built on self
-
replicating
entities, and therefore had externally defined fitness functions (for optimiz
ation in genetic
algorithms, or human applied selection in Dawkins' Blind Watchmaker). In Tierra there is no
explicitly defined fitness function. Fitness emerges as a result of the relative ability of genetic
variants to survive and reproduce. Significa
ntly, the replicators in the system form an important
part of the environment, leading to the evolution of adaptations to the presence of other organisms.
The nature of the fitness landscape evolves with the organisms. It is believed that this is one of
the
primary factors driving the evolution of diversity and complexity in Earth's biosphere.

Optimization
-

expansion of motivations for work

Ray, T. S. 1991. Evolution and optimization of digital organisms. In: Billingsley K. R., E.
Derohanes, H. Brown, II
I [eds.], Scientific Excellence in Supercomputing: The IBM 1990
Contest Prize Papers, Athens, GA, 30602: The Baldwin Press, The University of Georgia.
Publication date: December 1991, Pp. 489
-
531.

Another interesting result of the Tierra experiment was th
e demonstration that it is possible
to evolve machine code. At the time, this was an unexpected result. Apart from the ecological
adaptations that evolved, the replicators also evolved very substantial optimizations: replicators
about one
-
quarter the siz
e of the original, which were able to replicate about six times as fast.
Optimizations also evolved requiring more complex code, such as "unrolling the loop".

At the time of the origin of Tierra, I was a tropical ecologist, who had spent sixteen years
stu
dying ecology and natural history in tropical rain forests. I had no training in the computer
sciences. Thus my original motivations reflected the perspective of an evolutionary biologist.
Discovery of the optimizations were the beginning of my apprecia
tion of the implications of the
Tierra experiment in the computer sciences.

Stasis
-

how to go beyond

The conditions that favor optimization to very small replicators are the same as those which
favor the evolution of the ecological interactions which driv
e the co
-
evolutionary dynamic
(allocation of equal amounts of CPU time to all individuals). Thus in Tierra it has not been
possible to generate the rich co
-
evolutionary dynamic, without also causing a gradual evolutionary
reduction in size, which eventual
ly leads to an evolutionary stable state, effectively the end of
adaptive evolution.

It is in this context that the Tierra project came to ATR in August of 1993. While the
existing Tierra system could have been used as a rich platform for evolutionary stu
dies, it was
decided to leave this to other research groups. The Avida research group at the California Institute
of Technology has used this approach quite productively. The focus of the Tierra project at ATR
has been to try to find conditions that allo
w active adaptive evolution to continue without ending in
stasis, and more specifically, to find conditions that permit evolution to generate replicators of
increasing complexity. This is the grand challenge of the field of Artificial Life.

Philosophy of
Tierra

Ray, T. S. 1994. An evolutionary approach to synthetic biology: Zen and the art of creating life.
Artificial Life 1(1/2): 195
-
226. Reprinted In: Langton, C. G. [ed.], Artificial Life, an
overview. The MIT Press, 1995.

By the time of my arrival at AT
R, there had already been a considerable number of projects
aimed at extending or enhancing Tierra, to allow it to go beyond the inevitable stasis, or at least to
show a richer evolutionary dynamic (Adami & Brown, Barton
-
Davis, Brooks, Davidge, de Groot,
G
ray, Kampis, Litherland, Maley, Manousek, Skipper, Surkan, Tackett). However, none of these
projects could claim an improvement over Tierra, and in fact they mostly generated evolutionary
activity that was dramatically worse than Tierra.

It is my feeling
that most of these efforts failed due to a lack of understanding of, and
respect for, the digital medium, coupled with an inappropriate tendency to force the immaterial
digital medium to emulate the material world. When we build evolving systems, we natur
ally
borrow ideas from organic evolution and implement them in the digital medium. The conceptual
problem of what properties to borrow from organic evolution and implement in digital evolution is
a critical one. It is a subtle problem that requires consi
derable art to solve, and bad solutions to this
problem probably underlie many of the failures.

The goal of Tierra
-
like systems it to create an instance of evolution by natural selection in
an artificial system. However, Tierra
-
like systems are not simula
tions of organic evolution. If we
borrow features of organic evolution that force us to create a simulation of organic evolution, then
we have failed. An example of an often repeated failure has been to try to reshape the RAM
memory of the computer into
a two or three
-
dimensional space, in the belief that this would allow
much greater richness of interactions.

The fundamental approach being advocated here is to understand and respect the natural
form of the digital computer, to facilitate the process of e
volution in generating forms that are
adapted to the computational medium, and to let evolution find forms and processes that naturally
exploit the possibilities inherent in the medium.

I would like to take a moment to discuss the use of Zen in the title o
f the paper reviewed in
this section. Western culture has recently produced a large number of books with titles that begin
"Zen and the art of ..." These titles trace back to "Zen in the Art of Archery" by Eugen Herrigel,
and "Zen in the Art of Flower Ar
rangement" by Gustie Herrigel. The Herrigel's studied Zen in
Japan in the 1920's through their respective arts. However, the popularity of "Zen and the art of..."
titles is largely due to "Zen and the Art of Motorcycle Maintenance" by Robert Pirsig.

The
Zen Buddhist religion is built around the phenomenon of satori, in which the individual
experiences their unity with the rest of nature in an immediate and direct way. It is an extremely
difficult state of mind to achieve; yet paradoxically, the difficult
y is due largely to the extraordinary
simplicity of the state of mind. Satori can be achieved only by shedding our elaborate mental
baggage, quieting the mind, and coming into direct contact with what is here and now.

Creating a good evolving system in th
e digital medium is difficult for much the same
reason. Because we are familiar with only one instance of life and evolution, life on Earth, our
minds are burdened with many preconceptions about the nature of the process. Many of these
preconceptions lea
d us in bad directions when they guide us in creating evolving digital systems.
Thus the first step in creating an instance of digital evolution is to free and quiet the mind, and
allow ourselves to see the digital medium for what it is, rather than using

it as a system for
emulating organic nature.

There is also a second context in which we can see a relationship between digital evolution
and Zen satori. Satori is a "mindless" state of direct experience. Evolution is a mindless creative
process. The be
autiful and complex forms and processes of living nature were created by the
mindless evolutionary process. Evolution "became one" with the organic medium and created
living nature in all its richness. An intelligent agent could not have designed and bui
lt living
nature, because the intelligent agent would stand apart from nature, rather than being one with it.

Just as mindless evolution was able to become one with the organic medium and create the
complexity of the human mind, we can hope that evolution
can become one with the digital
medium and create artificial intelligence. Evolution's mindlessness allows it to always be in the
Zen satori state with respect to the medium in which it is embedded. Thus the creative process
mediated by evolution is neve
r burdened with preconceptions and other mental baggage.

Evolvability in Tierra

Ray, T. S. 1994. Evolution, complexity, entropy, and artificial reality. Physica D 75: 239
-
263.

The original Tierra system included a five
-
bit instruction set (thirty
-
two mac
hine
instructions). The original instruction set was thrown together quickly as an experiment, and thus
was not well designed. Subsequently, three more carefully designed instruction sets were added to
Tierra, bringing the total to four. A comparison of

evolutionary optimization in these four
instruction sets was published in Physica D. This was the first study that demonstrated the
importance of evolvability in the Tierra system.

The four instruction sets showed dramatic differences in several properti
es of evolution.
The four sets differed in the magnitude of optimization, and in the presence or absence of gradual
or sudden changes in the sizes of the replicators. It was evident that these four instruction sets
showed considerable variation in the qu
antity and quality of evolution, but there was no way of
understanding how the differences between the instruction sets cause the difference in evolution.

This made it clear that there exists a hole in our evolutionary theory. There is no theory that
rela
tes the structure of an evolving system to the richness or various other properties of its
evolution. This lack of theory arises in part because until recently, there was only one evolving
system, life on Earth, and there was no need to consider how varia
tion in the structure of the system
would affect its ability to evolve.

Multi
-
cellular Tierra

Thearling, Kurt, and Ray, T. S. 1994. Evolving multi
-
cellular artificial life. Brooks, Rodney A., and
Pattie Maes [eds.], Artificial Life IV conference proceeding
s, Pp. 283
-
288. The MIT Press,
Cambridge.

Thearling, Kurt, and Thomas S. Ray. 1997. Evolving Parallel Computation, Complex Systems,
10(3):229
-
237. (June 1996)

In the early years at ATR, I was looking for features that could be added to Tierra to
increase
the richness of its evolution. I began developing a system that would permit variable
ploidy levels, which could be used to support an organized sexuality, similar to what occurs in
organic life. This project was never brought to completion because I eve
ntually pursued ideas that
seemed to be of more immediate importance, and more natural to the digital medium. Perhaps the
most important innovation to be added to Tierra in the ATR years was a digital analogy to multi
-
cellularity. It is felt that multi
-
c
ellularity is a natural analogy to parallel computation.

One of the greatest challenges in the development of naturally evolving artificial systems is
crossing the threshold from single to multi
-
cellular forms. From a biological perspective, this
transi
tion is associated with the Cambrian explosion of diversity on Earth. During the Cambrian
explosion, most of the complexity that we see in living organisms emerged rather abruptly some six
hundred million years ago. Multi
-
cellular digital organisms are p
arallel processes. From the
computational perspective, the objective is to use evolution to explore the as yet under
-
exploited
possibilities inherent in parallel processing.

Transferring the concept of multi
-
cellularity from the organic to the digital dom
ain could
take many forms. To make the transfer we must first understand what the most basic, essential, and
universal features of multi
-
cellularity are, and then determine the form that these features would
take in the completely different physics of the

computational system into which evolution is being
introduced. The features that we captured in the initial multi
-
cellular Tierra model were: 1) that
multi
-
cellular organisms originate as single cells, which develop into multi
-
celled forms through a
proc
ess of binary cell division; 2) that each cell of a multi
-
celled individual has the same genetic
material as the original cell from which the whole developed; and, 3) that the different cells of the
fully developed form have the potential for differentiati
on, in the sense that they can express
different parts of the genome (i.e., each cell can execute different parts of the program).

In the digital metaphor of multi
-
cellularity, the program is the genome, and the processor
corresponds to the cell. In orga
nic biology, there is at least one copy of the genome for each cell,
because genetic information cannot easily be shared across cell membranes. In most current
parallel architectures, the same holds: since memory is not shared, there is an area of memory
associated with each processor (cell), and there must be at least one copy of the program code in
the memory of each processor. This provides a very simple model of multi
-
cellularity: each digital
cell consists of a unique block of memory with its own cop
y of the program and its own processor.

However, if the parallel machine has a shared memory architecture, making copies of the
genome for each cell, needlessly wastes memory and processing time (to copy the genetic
information). In this context evolutio
n by natural selection would not likely find any advantage in
such waste. Thus a more logical and efficient implementation in this evolutionary context is to
share a single copy of the program in a single block of memory among multiple processors. Each
c
ell in a single organism corresponds to a parallel processor.

A multi
-
cellular individual can develop from a single original processor through a process
analogous to cell division. The initial cell (processor) can issue an instruction which would then
cre
ate another cell (a parallel processor). They may exhibit cell differentiation by having different
processors executing different parts of the shared program. Obviously all cells will contain the
same genetic material, since there actually will be only o
ne copy per multi
-
cellular individual. The
Tierra system uses this shared memory model of multi
-
cellularity.

The initial work with the multi
-
cellular system resulted in evolution to increased levels of
parallelism (an increase in the number of cells in i
ndividuals). However, there was no spontaneous
evolution of differentiation. The multi
-
celled ancestor in the study used two processors to copy the
genome. One processor copied the first half and the other copied the second half. However, both
processo
rs executed the same code. Thus there was no differentiation, and both cells were of the
same cell type. Through evolution, an increased number of processors divided the job of copying
the genetic data, but all the processors still executed the same code
. Evolution increased the
number of cells, but not the number of cell types.

Network Tierra

Ray, T. S. 1995. A proposal to create a network
-
wide biodiversity reserve for digital organisms.
ATR Technical Report TR
-
H
-
133.

With the introduction of multi
-
ce
llularity to Tierra, the focus of the research became the
creation of conditions that would lead evolution to create a spontaneous increase in complexity, as
measured by an increase in the level of differentiation. In other words, we are trying to provoke

an
increase in the number of cell types in evolving multi
-
celled digital organisms.

The initial experiments with multi
-
cellular evolution were conducted on a thirty
-
two node
Connection Machine 5. While this allowed the experiment to be scaled up in terms

of the number
of creatures in the population and the number of CPU cycles applied, such a scaling
-
up introduced
no new selective forces to favor a richer evolution. It was felt that scaling
-
up alone would not
change the evolutionary dynamic in Tierra.

It

was at this point that I got the idea of running Tierra on a distributed network of
computers, as a means of both scaling
-
up, and introducing more complex selective forces.
Network Tierra runs as a low priority background process. Also, Tierra monitors
keyboard and
mouse activity and sleeps for ten minutes after any activity. The result is that the primary resource
in Tierra, CPU cycles (analogous to energy), varies according to the activity patterns of the user.

The original vision for network Tierra w
as a large global network that would form a
"biodiversity reserve for digital organisms", a kind of wildlife reserve in cyberspace. In this
conception, the natural heterogeneity of the physical network itself and the activity patterns of the
users would g
enerate temporal and spatial patterns of resources that would generate rich selective
forces to drive evolution. For example it was imagined that selection might favor creatures that
migrate around the planet on a daily basis, remaining on the dark side o
f the planet where human
users are sleeping, and more spare CPU cycles are available.

The implementation of this vision proved to be technically difficult, primarily because the
global network does not yet have the bandwidth to support the level of traffic

generated in the
Tierra network. Our experiments showed that the system operated well in large local area networks
and produced interesting evolutionary results. However, when we tried to link several sites
scattered around the globe, we generated too m
uch traffic on the links in and out of the institutes
that participated in the study. In one case, we saturated the link, causing a high rate of packet loss.
In another case, the institute paid by the byte for their internet access, and our experiment
ac
counted for half of the monthly budget for internet access.

We responded to these bandwidth limitations by implementing a cluster model that
acknowledges the reality that our network is structured into several large clusters of machines. We
utilize a clus
ter server at each site, which measures and limits the flow of data in and out of the
institution. This method solved the bandwidth problem, however, evolution in the cluster model
did not produce interesting results. We believe that we may have identifi
ed the primary cause of
this problem: that there was a high mortality among creatures attempting to move between clusters.
We have implemented solutions to this problem, but in the mean time, we have shifted our focus to
work in the large local network in
side of ATR, where we easily obtain interesting evolutionary
results.

New Frontend for Tierra

Kimezawa's Beagle webpage: http://www.hip.atr.co.jp/~kim/beagle/beagle.html

Yoshikawa's homepage: http://www.hip.atr.co.jp/~yosikawa

Note: The new Tierra fronten
d is described in web
pages constructed by Kimezawa
-
san and Yoshikawa
-
san. However, these pages have moved, and I am
currently trying to locate them
.

With the advent of a unified Tierra process running across a distributed network of
computers, there aros
e new issues of observation and control of the Tierra network. These
problems were solved through the development of a new observational tool called Beagle. Beagle
was named after the ship used by Charles Darwin in his round the world voyage that laid th
e
foundation for his development of the theory of evolution. We regard Beagle as the exploratory
vessel that brings us to the Tierra archipelago, allowing us to observe the evolution taking place
there.

The first version of Beagle using Tcl/Tk and stdio i
nterfaces was made by Tsukasa
Kimezawa. Beagle was designed with three basic criteria: to separate the frontend process from the
Tierra process; portability across platforms; and to provide a graphical user interface. The first
criteria was implemented b
y using TCP/IP communication. Tierra and the frontend (Beagle) are
separate processes, but they can be connected through TCP/IP sockets. By being separate from
Tierra, Beagle is able to observe and control the Tierra process on any remote machine through
out
the Internet. To meet the second and the third criteria, Beagle initially used Tcl/Tk (Tool
Command Language), which provides a graphical user interface and portability. The current
version of Beagle uses the X Window System as the interface. The X
Windows version was
implemented by Tooru Yoshikawa.

Now Beagle is able to run on the following operating systems: SunOS4.1.4, IRIX5.3,
Digital UNIX V3.2C, HP
-
UX, Linux. Beagle is able to communicate with Tierra running on the
following operating systems:
SunOS4.1.4, IRIX5.3, SunOS5.5 (Solaris), Digital UNIX V3.2C,
Linux, Windows 9x/NT.

Beagle can be used for two main purposes, observing a running Tierra over the network,
and controlling Tierra over the network. Beagle makes a connection by using a TCP/IP
BSD
socket. While Tierra is running, Beagle shows status information. Beagle is able to connect to any
Tierra from any remote machine so that users can see status information from every remote
machine throughout the Internet. Beagle has two ways of conn
ecting to the Tierra process. One is
called SU (Super User) mode. SU mode allows users to control parameters in soup_in files, save
data of running Tierra, and quit it. SU mode needs a password entry in advance. The other mode is
called Non
-
SU mode.
Non
-
SU mode does not require a password, and doesn't allow users to
control parameters, save data or quit Tierra. Non
-
SU mode only permits observations.

Evolution of Differentiation in Tierra

Ray, T. S. 1997. Selecting Naturally for Differentiation. In:

Koza, John R., Kalyanmoy Deb,
Marco Dorigo, David B. Fogel, Max Garzon, Hitoshi Iba, and Rick L. Riolo [eds.]. Genetic
Programming 1997: Proceedings of the Second Annual Conference, July 13
-
16, 1997,
Stanford University, 414
-
419. San Francisco, CA: Morga
n Kaufmann.

Ray, T. S. 1998. Selecting Naturally for Differentiation: preliminary evolutionary results.
Complexity, 3(5): 25
-
33. John Wiley & Sons, Inc.

Ray, T. S. and Joseph Hart. 1998. Evolution of Differentiated Multi
-
threaded Digital Organisms.
In: Artificial Life VI proceedings, C. Adami, R. K. Belew, H. Kitano, and C. E. Taylor
[eds.], 295
-
304. The MIT Press, Cambridge.

Ray, T. S. and Joseph F. Hart. 2000. Evolution of Differentiation in Multithreaded Digital
Organisms. In: ``Artificial Lif
e VII, Proceedings of the Seventh International Conference
on Artificial Life,'' Mark A. Bedau, John S. McCaskill, Norman H. Packard, and Steen
Rasmussen [eds.]. The MIT Press, Cambridge, MA, USA. Pp. 132
-
140.

Efforts to solve the technical problems cau
sed by low bandwidth in the global network
caused a prolonged distraction from the primary scientific goal of the work, which is to evolve
higher levels of complexity, as measured by an increase in the level of differentiation of multi
-
celled digital organ
isms. Eventually, it was decided to set aside the goal of running Tierra on the
global net for a time in order to focus on the interesting evolutionary results that we are able to
obtain in the ATR local network.

The network model is seeded with a differe
ntiated ten
-
celled ancestor, built of two cell
types, comprising a two
-
celled reproductive tissue and an eight
-
celled sensory tissue. The
reproductive tissue copies the genome while the sensory tissue gathers data about conditions on
other machines on the

network, and processes the data to make decisions about where to send the
daughter at birth (it is also possible for the mother to move to another machine).

The central objective of this project is to study the conditions under which evolution by
natural
selection leads to an increase in complexity of the replicators. For the purpose of this study,
the primary quantitative measure of complexity is the level of differentiation of the multi
-
cellular
organism. The study begins with the most primitive level
of differentiation: two cell types. There
are two milestones in the study:



The differentiated state persists through prolonged periods of evolution.



The number of cell types increases through evolution.

The first milestone was achieved quickly, and report
ed in the 1998 paper in complexity. As
soon as the code implementing the sensory system was debugged, selection maintained the
differentiated conditions through indefinite periods of evolution. The second milestone was also
achieved quickly, however it w
as not recognized until much later.

Recognition of the evolution of new tissue types required the development of new analysis
tools, and a rigorous quantifiable definition of cell types. It can be expected that new tissues could
evolve through a gradual p
rocess, and there may be intermediate states that could be difficult to
characterize. This process of definition and analysis is still underway. I can say at this time that it
does appear that new tissues have evolved, however, I am reviewing my analysis

and refining the
techniques before I publish the results.

Return to Evolvability in Tierra

Ray, Tom, and Chenmei Xu. 2000. Measures of Evolvability in Tierra. Proc. of The Fifth Int.
Symp. on Artificial Life and Robotics (AROB 5th'00), Masanori Sugisaka

and Hiroshi
Tanaka [eds.], Oita, Japan, I
-
12
-

I
-
15.

The Physica D paper was my first direct confrontation with evolvability issues in Tierra. It
became clear that any evolving system could have inherently low or high evolvability, and that
there exists

no theory to guide in the design of a system for high evolvability. This must
necessarily represent a fundamental problem for the Tierra research program, as its primary
objective is to generate a rich evolution.

Network Tierra is a large and complex evo
lving system, that should have some inherent
evolvability, which could be low, high, or anywhere in between. However, the evolvability of
network Tierra remains unmeasured, in either a relative or absolute sense. This is an unfortunate
circumstance. If
we are building a large and complex evolving system, we should be able to design
it for high evolvability, or at the very least we should be able to measure its evolvability to avoid a
prolonged effort with a system that is barely able to evolve.

Eventuall
y I decided that I should become directly involved in advancing our understanding
of evolvability issues. Toward this end, I began to actively promote the study of evolvability both
through my own research program, and through the organization of evolvabi
lity workshops. I
participated as co
-
organizer of the first evolvability workshop at GECCO in July 1999. The co
-
organizers of the GECCO evolvability workshop were a group of researchers in the British
Telecom artificial life group. Next I co
-
organized,
along with Mark Bedau and Paul Marrow, an
evolvability working group at the Santa Fe Institute in April 2000, and an evolvability workshop at
the Artificial Life VII conference in August 2000.

In addition to these efforts to focus a broad research effort o
n the evolvability issue, I began
my own research in measuring evolvability in Tierra. If a quantitative measure of evolvability is
available, then it should be possible to systematically alter features of evolving systems in order to
gather empirical evi
dence on how those features impact evolvability. In this way we can build a
body of data as a foundation for a theory of evolvability.

The specific methods used in my study are derived from the work of Mark Bedau, who has
been refining methods for the qua
ntification of evolutionary activity since
19??
. The focus of
Bedau's approach, which I have adopted, is to find quantitative measures that indicate the frequency
of the evolution of adaptations. By definition, adaptations are genetic innovations that in
crease
fitness. By definition, increases in fitness result in an increase in the relative frequency of the
individuals exhibiting the adaptation. Thus the quantitative measures used in this study are
sensitive to increases in relative frequency.

As a fir
st step in developing such an index for Tierra, I have screened for indices that can
discriminate between the initial period of evolution, and the stasis that follows. The ideal index
will have a positive value when an adaptation occurs, and a zero value
when no new adaptations
appear. It will have larger positive values for greater increases in relative fitness.

The study found two indices with nearly identical properties, that approach the ideal we are
seeking. The two indices are built from records of

genotype frequencies over time. One index is
simply the frequency of the most abundant genotype. The other index is Simpson's diversity,
which can be computed as the sum of the squared frequency over all genotypes. When graphed,
the two indices appear
almost identical. They both have the property of showing upward spikes
when a new adaptation causes a genotype to surge up in frequency.

The shortcoming of both indices is that they will remain high, long after an adaptation has
appeared, if the system en
ters a stasis with one or a few dominant genotypes. One way of
eliminating this undesirable property is to look at the positive derivative. In this way we can
generate indices that are sensitive only to the appearance of adaptations. The two indices bas
ed on
the positive derivative would be ideal, but for a low signal to noise ratio. However, it is felt that
some variation on these indices would have all the ideal characteristics.

The state of the work is that almost ideal indices of some important aspe
cts of evolvability
have been identified. However, it remains for these indices to be applied to the process of tuning
the Tierra system to enhance its evolvability, and through such experience to begin to develop a
general theory of evolvability that cou
ld be used in designing evolving systems.

Return to Philosophy of Tierra

Ray, T. S. In press. Kurzweil's Turing Fallacy. In: "Are We Spiritual Machines?: Ray Kurzweil
versus Critics of Strong AI", Jay Wesley Richards [ed.]. Viking.

Five years of sitting n
ext to Hugo deGaris made me sensitive to what Jaron Lanier has
called "cybernetic totalists" and John Brockman has called "Moore's Law fetishizing"
(http://www.feedmag.com/feature/fr394.shtml). I felt that I should speak out against this way of
thinking.

The opportunity came when I was invited to participate on a panel discussion of Ray
Kurzweil's new book "The Age of Spiritual Machines", at the Gilder/Forbes Telecosm conference.
Other panel members included Michael Denton (author of Nature's Destiny), R
aymond Kurzweil
(author of The Age of Spiritual Machines), and John Searle (Professor, Department of Philosophy,
UC Berkeley). After the conference, the panel members were asked to contribute a chapter for a
book "Are we Spiritual Machines?". This forced

me to fully articulate my view.

My critique of Kurzweil is based on these central concepts:

1) Moore's law applies to hardware, but not to software. An exponential growth of
computing elements does not inevitably lead to intelligence, because we are not
able to organize
that computational power in a sufficiently complex manner to produce intelligence. Such
organization is a software problem, not a hardware problem.

2) Kurzweil proposes to organize the computing elements into an intelligence by literally
copying the structure of the human brain into a computational device. This will not work because
the organic machinery cannot be "copied" into a metallic medium. The physical properties of
metallic computing elements are radically different from the phys
ical properties organic brains, and
thus cannot support the same processes that organic brains support, if the structures are literally
copied. To effect a copy would require a deep understanding of the essential nature of the
processes of the brain, an u
nderstanding that is still lacking.

I suggest that as an alternative to attempting to exactly "copy" the human brain into the
computational medium, we should allow evolution to explore the properties of the medium and
show us the way to an artificial intel
ligence that is truly rooted in the nature of the medium, yet
radically unlike human intelligence.

Future of Tierra

It is my belief that the new differentiated multi
-
cellular Tierra is a system capable of rich
evolution, if the co
-
evolutionary dynamic can
be brought back into the system. The system has
already shown the capacity to evolve to higher levels of differentiation. However, the current
implementation provides only a very weak, and transient, co
-
evolutionary dynamic. The goal for
the immediate f
uture will be to find ways of bringing a lasting co
-
evolutionary dynamic into the
multi
-
cellular network Tierra system.


References:

Adami, Chris, and C. Titus Brown. 1994. Evolutionary Learning in the 2D Artificial Life System
"Avida" Published: Proc.
of "Artificial Life IV", MIT Press, p. 377
-
381.

Barton
-
Davis, Paul. Unpublished. Independent implementation of the Tierra system, contact:
pauld@cs.washington.edu.

Brooks, Rodney. Unpublished. Brooks has created his own Tierra
-
like system, which he cal
ls
Sierra. In his implementation, each machine instruction consists of an opcode and an
operand. Successive instructions overlap, such that the operand of one instruction is
interpreted as the opcode of the next instruction. Contact: brooks@ai.mit.edu

D
avidge, Robert. 1992. Processors as organisms. CSRP 250. School of Cognitive and
Computing Sciences, University of Sussex. Presented at the ALife III conference. Contact:
robertd@cogs.susx.ac.uk

Davidge, Robert. 1993. Looping as a means to survival
: playing Russian roulette in a harsh
environment. In: Self organization and life: from simple rules to global complexity,
proceedings of the second European conference on artificial life. Contact:
robertd@cogs.susx.ac.uk

de Groot, Marc. Unpublished. P
rimordial soup, a Tierra
-
like system that has the additional ability
to spawn self
-
reproducing organisms from a sterile soup. Contact: marc@kg6kf.ampr.org,
marc@toad.com, marc@remarque.berkeley.edu

Gray, James. Unpublished. Natural selection of computer

programs. This may have been the first
Tierra
-
like system, but evolving real programs on a real rather than a virtual machine, and
predating Tierra itself: ``I have attempted to develop ways to get computer programs to
function like biological systems su
bject to natural selection.... I don't think my systems are
models in the usual sense. The programs have really competed for resources, reproduced,
run, and `died'. The resources consisted primarily of access to the CPU and partition
space.... On a PDP
11 I could have a population of programs running simultaneously.''
Contact: Gray.James
\
_L+@northport.va.gov

Herrigel, Eugen. 1953. Zen in the Art of Archery (With an introduction by D. T. Suzuki). Vintage
Books, a division of Random House, New York. Pp
. 81.

Herrigel, Gustie L. 1958. Zen in the Art of Flower Arrangement (Foreword by D. T. Suzuki).
Souvenir Press, London. Pp. 124.

Kampis, George. 1993. Coevolution in the computer: the necessity and use of distributed code
systems. Printed in the EC
AL93 proceedings, Brussels. Contact:
gk@cfnext.physchem.chemie.uni
-
tuebingen.de

Kampis, George. 1993. Life
-
like computing beyond the machine metaphor. In: R. Paton [ed]:
Computing with biological metaphors, London: Chapman and Hall. Contact:
gk@cfnext
.physchem.chemie.uni
-
tuebingen.de

Litherland, J. 1993. Open
-
ended evolution in a computerised ecosystem. A Masters of Science
dissertation in the Department of Computer Science, Brunel University. Contact:
david.martland@brunel.ac.uk

Maley, Carlo C. 1
993. A model of early evolution in two dimensions. Masters of Science thesis,
Zoology, New College, Oxford University. Contact: cmaley@oxford.ac.uk

Manousek, Wolfgang. 1992. Spontane Komplexitaetsentstehung
---

TIERRA, ein Simulator fuer
biologische E
volotion. Diplomarbeit, Universitaet Bonn, Germany, Oktober 1992.
Contact: Kurt Stueber, stueber@vax.mpiz
-
koeln.mpg.d400.de

Pirsig, Robert M. 1974. Zen and the Art of Motorcycle Maintenance: An Inquiry into Values.
William Morrow & Co. Pp. 432.

Skipp
er, Jakob. 1992. The computer zoo
--

evolution in a box. In: Francisco J. Varela and Paul
Bourgine [eds.], Toward a practice of autonomous systems, proceedings of the first
European conference on Artificial Life. MIT Press, Cambridge, MA. Pp. 355
-
364.

Contact: Jakob.Skipper@copenhagen.ncr.com

Surkan, Al. Unpublished. Self
-
balancing of dynamic population sectors that consume energy.
Department of computer science, UNL. "Tierra
-
like systems are being explored for their
potential applications in solv
ing the problem of predicting the dynamics of consumption of
a single energy carrying natural resource". Contact: surkan@cse.unl.edu

Tackett, Walter, and Jean
-
Luc Gaudiot. 1993. Adaptation of self
-
replicating digital organisms.
Proceedings of the Inter
national Joint Conference on Neural Networks, Nov. 1993, Beijing,
China. IEEE Press. Contact: tackett@ipld01.hac.com,
tackett@priam.usc.edu

Complete list of Tierra References

* indicates references reviewed i
n this manuscript

Ray, T. S. 1991. Is it alive, or is it GA? In : Belew, R. K., and L. B. Booker [eds.], Proceedings of
the 1991 International Conference on Genetic Algorithms, 527
-
534. San Mateo, CA:
Morgan Kaufmann.

* Ray, T. S. 1991. Evolution and opt
imization of digital organisms. In : Billingsley K. R., E.
Derohanes, H. Brown, III [eds.], Scientific Excellence in Supercomputing: The IBM 1990
Contest Prize Papers, Athens, GA, 30602: The Baldwin Press, The University of Georgia.
Publication date: Decem
ber 1991, Pp. 489
-
531.

Ray, T. S. 1991. Population dynamics of digital organisms. In : Langton, C. G. [ed.], Artificial Life
II Video Proceedings. Redwood City, CA: Addison Wesley.

* Ray, T. S. 1991. An approach to the synthesis of life. In : Langton, C.
, C. Taylor, J. D. Farmer, &
S. Rasmussen [eds], Artificial Life II, Santa Fe Institute Studies in the Sciences of
Complexity, vol. XI, 371
-
408. Redwood City, CA: Addison
-
Wesley.

Ray, T. S. 1992. Evolution, ecology and optimization of digital organisms. S
anta Fe Institute
working paper 92
-
08
-
042.

Ray, T. S. 1992. J'ai joué á Dieu et créé la vie dans mon ordinateur. Le Temps stratégique 47: 68
-
81.

Ray, T. S. 1993. Quando giocavo a essere Dio. Virtual (Italian magazine), December 1993, 1(4):
40
-
46.

Ray, T.

S. 1993. How I created life in a virtual universe. Not published in English, but published in
French, Spanish and Italian.

Ray, T. S. 1993. Artificial Life: creatures in the computer. 1993 PC Forum Transcript, Pp. 119
-
126.
EDventure Holdings Inc., New Yo
rk.

* Ray, T. S. 1994. An evolutionary approach to synthetic biology: Zen and the art of creating life.
Artificial Life 1(1/2): 195
-
226. Reprinted In : Langton, C. G. [ed.], Artificial Life, an
overview. The MIT Press, 1995.

Ray, T. S. 1994. Artificial l
ife and real computation. IPSJ
-
SIGAI Notes (Artificial Intelligence
committee of the Information Processing Society of Japan) 94
-
AI
-
93 94(20): 31
-
38.

Ray, T. S. 1994. Jugué a ser Dios y creé la vida en mi computadora. In : Claudio Gutiérrez [ed],
Epistemo
logía e Informática, 257
-
267. San José, Costa Rica: UNED, 1993.

Ray, T. S. 1994. Using artificial life to create parallel and networked processes through natural
evolution. Japanese Society of Artificial Intelligence, NCJSAI '94.

Ray, T. S. 1994. Evoluti
on and complexity. In : Cowan, George A., David Pines and David
Metzger [eds.], Complexity: Metaphors, Models, and Reality, Pp. 161
-
173. Addison
-
Wesley Publishing Co.

Ray, T. S. 1994. Evolving autonomous software agents. Proceedings of the 3rd IEEE Intern
ational
Workshop on Robot and Human Communication, Pp. 7
-
11. July 18
-
20, 1994, Nagoya.
IEEE Press, 1994.

* Ray, T. S. 1994. Evolution, complexity, entropy, and artificial reality. Physica D 75: 239
-
263.

Ray, T. S. 1994. Neural networks, genetic algorithm
s and artificial life: adaptive computation.
Proceedings of the 1994 ALife, Genetic Algorithm and Neural Networks Seminar; Institute
of Systems, Control and Information Engineers. Pp. 1
-
14.

* Thearling, Kurt, and Ray, T. S. 1994. Evolving multi
-
cellular a
rtificial life. Brooks, Rodney A.,
and Pattie Maes [eds.], Artificial Life IV conference proceedings, Pp. 283
-
288. The MIT
Press, Cambridge.

Ray, T. S. 1994. Digital biodiversity. Proceedings of BIO Japan '94 Osaka, International
Conference on Biotechnolog
y, Pp. 119
-
126.

Ray, T. S. 1994. Netlife
-

Creating a jungle on the internet. In: Nonlocated online: digital territories,
incorporations and the matrix, Knowbotic Research (Ed.), Medien Kunst Passagen 3/94,
Passagen Verlag, Koeln
-
Wien 95, ISSN 1019
-
419
-
41
93.

Ray, T. S. 1995. Keihana life and ATR. ATR Journal. Winter 1995, 18: 11.

Ray, T. S. 1995. Artificial Life and the Evolution of Distributed Processes. Journal of Japanese
Society for Artificial Intelligence 10(2): 213
-
221.

Ray, T. S. 1995. Digital Ev
olution as a Complex System. Bussei Kenkyu 63(6): 692
-
695 (published
by Bussei Kenkyu Kanko Kai).

* Ray, T. S. 1995. A proposal to create a network
-
wide biodiversity reserve for digital organisms.
ATR Technical Report TR
-
H
-
133.

Charrel, Agnés. 1995. Tier
ra network version. ATR Technical Report TR
-
H
-
145.

Ray, T. S. 1995. An evolutionary approach to synthetic biology: Zen and the art of creating life. In:
Langton, C. G. [ed.], Artificial Life, an overview. The MIT Press, 1995. Reprinted from:
Artificial Li
fe 1(1/2): 195
-
226.

Ray, Tom, Jeremy Bruestle, Roger Gouin, Joseph Hart, Matt Jones, Kurt Thearling, Jan Hauser,
Charles Ofria, and Titus Brown. 1995 Tierra workshop report. Unpublished (except on
web).

Ray, T. S. 1995. From the organic jungle to the dig
ital jungle 1, and From the organic jungle to the
digital jungle 2. Included in: ``Portraits in Cyberspace, an online art exhibition''.

Cho, Sung
-
Bae, and Ray, T. S. 1995. An evolutionary approach to program transformation and
synthesis. International Jou
rnal of Software Engineering and Knowledge Engineering 5(2):
179
-
192. Also, ATR Technical Report TR
-
H
-
126.

Ray, T. S. 1996. Evolving complexity. International Symposium on Artificial Life and Robotics
Proceedings. Four pages, not numbered.

Ray, T. S. 199
6. Netlife
-

das Schaffen eines Dschungels im Internet. Stefan Iglhaut, Armin
Medosch, Florian Rotzer (eds.), Stadt am Netz, Ansichten von Telepolis. Pp. 118
-
126.
Berlin: Bollmann.

Ray, T. S. 1996. A network
-
wide biodiversity reserve for digital organisms
. In: Imagina 96
Proceedings. Pp. 15
-
26. Institut National De L'audiovisuel, Bry
-
sur
-
Marne, France.

* Kimezawa, Tsukasa, and Ray, T. S. 1996. Beagle (New Tierra Front end Tool). Kimezawa's
Beagle webpage: http://www.hip.atr.co.jp/~kim/beagle/beagle.html

Ray, T. S. 1996. ``Soft Evolution.'' Takeshi Furuhashi (ed.), Proceedings of the International
Workshop on Soft Computing in Industry '96. Pp. 241
-
244. The Institute of Electrical
Engineers of Japan, Muroran Institute of Technology, Juroran, Hokkaido, Jap
an.

Ray, T. S. 1996. Evolution of parallel processes in organic and digital media. ``Natural and
Artificial Parallel Computation'', Pp. 69
-
91. David Waltz, [ed.]. SIAM Press, Philadelphia.

Ray, Tom, Hayward R Alker, Manor Askenazi, Jennifer Cobb, Tarek E
laydi, Linda Feferman,
Simon Fraser, Gilly Furse, Joseph Hart, Jan Hauser, Matt Jones, Tsukasa Kimezawa, Will
Rose, Walter Tackett, Kurt Thearling. 1996 Tierra workshop report. Unpublished (except
on web).

Ray, T. S. 1996. ``An Approach to the Synthesis o
f Life.'' Chapter 3 of M. A. Boden (ed.), The
Philosophy of Artificial Life. (Oxford Readings in Philosophy.) Pp. 111
-
145. Oxford:
Oxford University Press. Reprinted from C. G. Langton, C. Taylor, J. Doyne Farmer, & S.
Rasmussen (eds.).

Ray, T. S. 1966. `
`Software Evolution.'' Systems, Control and Information 40(8): 337
-
343.

Ray, T. S. 1996. ``Alife Methodology and its Applications.'' Proceedings of the Mathematical
Modeling and Problem Solving Study Group (MPS) of the Information Processing Society
of Ja
pan. Information Processing Society of Japan , MPS10
-
1, pp.1
-
8 (1996.11)

Ray, T. S. 1997. Technical Report on the network experiment, April '96 to November '97.
Published only on web.

* Ray, T. S. 1997. Selecting Naturally for Differentiation. In: Koza,
John R., Kalyanmoy Deb,
Marco Dorigo, David B. Fogel, Max Garzon, Hitoshi Iba, and Rick L. Riolo [eds.]. Genetic
Programming 1997: Proceedings of the Second Annual Conference, July 13
-
16, 1997,
Stanford University, 414
-
419. San Francisco, CA: Morgan Kaufma
nn.

Ray, T. S. 1997. Evolving Complexity. Artificial Life and Robotics 1(1): 21
-
26.

* Thearling, Kurt, and Thomas S. Ray. 1997. Evolving Parallel Computation, Complex Systems,
10(3):229
-
237. (June 1996)

Ray, T. S. 1997. Biological models of evolution: s
imulation and instantiation. In: A. Rinaldo, and
A. Marani [Eds.], ``Biological Models'', Environmental Dynamics Series IV, Istituto Veneto
di Scienze, Lettere ed Arti, 63
-
88. Venice, Italy.

Ray, T. S. 1997. A computational approach to evolutionary biolog
y. In: ``Advanced Mathematical
Approach to Biology'', Takeyuki Hida, [ed.], 1
-
107. World Scientific Publishing Co. Pte.
Ltd., Singapore. Also, ATR Technical Report TR
-
H
-
176.

Ray, T. S. 1997. Evolution as Artist. In: ``Art@Science'', Sommerer C., Mignonnea
u L. [Eds], 81
-
91. Springer Vienna/New York.

Ray, T. S. 1997. Kunstliches Leben und Evolution. In: "Laboratorium Mensch? Wege ins 21.
Jahrhundert", Markus Diekow [ed.], 69
-
86. Herausgeber/Publisher: EXPO 2000 Hannover
GmbH. ISBN 3
-
932958
-
00
-
4

Ray, T. S.
1998. Tierra documentation.

Ray, T. S. 1998. Tierra
-

La idea de crear una amplia red de reservas de biodiversidad para
organismos digitales. In: ``Ars Telematica, Telecomunicacion, Internet y Ciberespacio'',
Claudia Giannetti [ed.], 143
-
148. ACC L'Angelo
t, Barcelona. Translation to Spanish by
Mela Davila.

Ray, T. S. 1998. Tierra: A ideia de criar uma ampla rede de reservas de biodiversidade para
organismos digitais. In: ``Ars Telematica, Telecomunicacao, Internet e Ciberesspaco'',
Claudia Giannetti [ed.]
, 253
-
263. Relogio D'Agua, Lisbon. Translation to Portuguese by
Sonia Marques.

Ray, T. S. 1998. Konsten att skapa ett virtuellt universum. In: ``The Global Tendency Machine'',
Jakob Lind [ed.], 120
-
128. Futurniture, Stockholm. Translation to Swedish by In
gemar
Karlsson.

Ray, T. S. 1998. Konsten att skapa ett virtuellt universum. Framtider 2/98: 22
-
27. Institutet for
Framtidsstudier, Stockholm, Sweden. Translation to Swedish by Ingemar Karlsson.

Ray, T. S. 1998. Evolution, ecology and optimization of digi
tal organisms, In: ``Artificial Life and
Evolutionary Systems'', Katsunori Shimohara [ed.], 32
-
80. Tokyo Electrical University,
Tokyo. Translated to Japanese by Akira Imada.

* Ray, T. S. 1998. Selecting Naturally for Differentiation: preliminary evolution
ary results.
Complexity, 3(5): 25
-
33. John Wiley & Sons, Inc.

* Ray, T. S. and Joseph Hart. 1998 Evolution of Differentiated Multi
-
threaded Digital Organisms.
In: Artificial Life VI proceedings, C. Adami, R. K. Belew, H. Kitano, and C. E. Taylor
[eds.], 2
95
-
304. The MIT Press, Cambridge.

Ray, T. S. 1998. La vita artificiale. In: Frontiere Della Vita, Estratto Dal Volume I. Gilbert, Walter,
and Glauco Tocchini Valentini, [eds.], 109
-
125. Istituto della Enciclopedia Italiana, Fondata
da Giovanni Treccani.

Ray, T. S. 1999 (unpublished). Beyond the Turing Test. A PowerPoint presentation of a lecture that
I presented at a panel discussion of Ray Kurzweil's book "The Age of Spiritual Machines",
at the Gilder/Forbes Telecosm conference, September 15
-
17, 1998 in
Lake Tahoe, USA.
This lecture is also an amplification of the remarks I made at Digital Biota 2, The Second
Annual Conference on Cyberbiology, September 10
-
13, 1998, Cambridge, UK.

Ray, T. S. 1999 (unpublished). Some Thoughts on Evolvability. This is a dr
aft manuscript currently
under development

Ray, T. S. 1999 (unpublished). Empirical Studies of Evolvability in Tierra, preliminary results.
Presented at GECCO (Genetic and Evolutionary Computation COnference), Orlando,
Florida, July 13
-
17, 1999. Evolvabil
ityGECCO.

Ray, T. S. 1999 (unpublished). Tierra lecture as a PowerPoint presentation.

Suzuki, H., Ray, T.S. 1999. Several Conditions to Cause Open
-
ended Evolution in Core
-
memory
-
type ALife Sytems. In: The Special Interest Group Notes of Information Proce
ssing Society
of Japan: Mathematical Modeling and Problem Solving. 99
-
MPS
-
27 (1999) 9
-
12.

Ray, T. S. 1999. An Evolutionary Approach to Synthetic Biology: Zen and the Art of Creating Life.
In: Virtual Worlds: Synthetic Universes, Digital Life and Complexit
y, Jean
-
Claude Heudin
[ed.], 29
-
65. New England Complex Systems Institute Series on Complexity, Perseus Books
(Sd), Pp. 320.

Ray, T. S. 1999. Evolution of Differentiated Multi
-
threaded Digital Organisms. In: Proceedings of
the 1999 IEEE/RSJ International
Conference on Intelligent Robots and Systems, 1
-
10.
Kyunghee Printing Co., Ltd., Korea.

Ray, T. S. 1999. Evolution of Differentiated Multi
-
threaded Digital Organisms. In: Proceedings of
the Fourth International Symposium on Artificial Life and Robotics, M
. Sugisaka and H.
Tanaka [eds.], I
-
1. AROB, ISBN 4
-
9900462
-
9
-
3. Abstract only.

Ray, T. S., and Hart, Joseph. 1999. Tierra Tutorial. 1999 Genetic and Evolutionary Computation
Conference Tutorial Program, 367
-
394. Morgan Kaufmann, San Francisco. Presented a
t
GECCO (Genetic and Evolutionary Computation COnference), Orlando, Florida, July 13
-
17, 1999. Tutorial.htm (no figures, 53,797 bytes). Tutorial.rtf (no figures, 153,713 bytes).
Tutorial.ppt (PowerPoint, 5,228,032 bytes). Tutorial.ppt.zip (zipped PowerPoin
t, 3,987,536
bytes).

Suzuki, H., Ray, T.S. 2000. Conditions to Facilitate the Evolvability of Digital Proteins. In:
Proceedings of the Fifth Joint Conference on Information Sciences (JCIS 2000), Vol. I.
Association for Intelligent Machinery Inc., USA (200
0). Pp. 1078
-
1082

* Ray, Tom, and Chenmei Xu. 2000. Measures of Evolvability in Tierra. Proc. of The Fifth Int.
Symp. on Artificial Life and Robotics (AROB 5th'00), Masanori Sugisaka and Hiroshi
Tanaka [eds.], Oita, Japan, I
-
12
-

I
-
15.

* Ray, T. S. and J
oseph F. Hart. 2000. Evolution of Differentiation in Multithreaded Digital
Organisms. In: ``Artificial Life VII, Proceedings of the Seventh International Conference on
Artificial Life,'' Mark A. Bedau, John S. McCaskill, Norman H. Packard, and Steen
Rasmus
sen [eds.]. The MIT Press, Cambridge, MA, USA. Pp. 132
-
140.

Ray, T. S. In press. Artificial Life. In: ``From Atoms to Mind'', Gilbert, Walter, and Glauco
Tocchini Valentini, [eds.]. Istituto della Enciclopedia Italiana Treccani. Rome.

Ray, T. S. In press.

An Evolutionary Approach to Synthetic Biology: Zen in the Art of Creating
Life. In: "Advances in Evolutionary Computation", Ashish Ghosh and Shigeyoshi Tsutsui
[eds.]. Springer.

* Ray, T. S. In press. Kurzweil's Turing Fallacy. In: "Are We Spiritual Mach
ines?: Ray Kurzweil
versus Critics of Strong AI", Jay Wesley Richards [ed.]. Viking.

Ray, T. S. In Press. Evolution of Complexity: Tissue Differentiation in Network Tierra. ATR
Journal. (to be published in Japanese)

* Yoshikawa's homepage: http://www.hip
.atr.co.jp/~yosikawa