Artificial
Life IX: Proceedings of the Ninth International
Conference on the Simulation and Synthesis
of Living Systems
by Jordan Pollack,
Mark Bedau, Phil Husbands, Takashi Ikegami
and Richard A. Watson, Eds.
The MIT Press,
Cambridge, Massachusetts, 2004
600 pp., illus. Paper, $80.00
ISBN: 0-262-66183-7
Reviewed
by Stefaan Van Ryssen
Hogeschool Gent
stefaan.vanryssen@pandora.be
Artificial Life
(AL) has never been and probably never
will be the art of re-creating life in
a test tube, even if the tabloids and
some conservative lobbies want the public
to believe it is, in the hope of selling
more copies or winning more souls. Instead
AL isin the words of editor
Jordan B. Pollack"the study
of the organizational principles of life,
rather than the study of carbon-based
life as it exists on Earth. Some of the
fundamental questions of AL are:
- What are the principles
of evolution, learning and growth that
can be understood well enough to simulate
as an information process?
- Can robots be built
faster and cheaper by mimicking biology
than by the product design process used
for automobiles and airplanes?
- What kind of constraints
should be placed on sciences, such as
"Wet ALife," which work with self-replicating
elements?
- What components
of physics and chemistry support emergence
and automatic discovery of physical
and cognitive mechanisms of life forms?
- How can we unify
theories from dynamical systems, game
theory, evolution, computing, geophysics,
and cognition?" (p. xii)
This certainly is
a tall order, but research done by the
scientists who attended the Ninth International
Conference on the Simulation and Synthesis
of Artificial Life in Boston in 2004 testifies
to its feasibilityin the long
run. No less than 97 papers were selected
for presentation and published in the
proceedings.
It is hard to find any trends or even
highlights in this highly entropic collection.
The papers organized themselves alphabetically
on author's names in nine subfields, so
it is not surprising that the subfield
of Self-Organization has caught the eye
of the popular scientific pressalso
because it delivers some spectacular results
like self-replicating machines (Efstathios
Mytilinaios et al).
The subfields of Robotic Studies, Formal
Games and Automata, Artificial Chemistries
and Evolutionary and Adaptive Dynamics
evidently are blooming, since they take
up more than half of the contributions
but it was in 'Language, Brain, Culture'
and 'Art and Philosophy' that I could
find the least technical though certainly
not always the most accessible material.
'Language' in this context is not to be
taken as the highly sophisticated communications
systems we colloquially use but rather
as a set of simple rules and a few words
the evolution of which is studied in a
computer-simulated environment. And this
says it all. Artificial Life studies nowadays
consist mostly of mathematical and statistical
analysis of emergent properties of very
simple computer-dwelling entities, fitted
with a neural network and evolving by
means of a genetic algorithm in a highly
simplified environment. The majority of
the papers here do include just these
elements: evolution, emergent behaviour
or emergent properties and contextual
or environmental restraints. From these
principles, that are emulated in very
different ways, very divergent effects
can be seen to arise: from the expression
of desire to the recognition of the shape
of objects to the emergence of grammar.
Being the proceedings of a conference,
most papers are highly technical and require
a good understanding of the typical language
of neural networks and genetic algorithms,
and a background in at least the basics
of Artificial Life or the older science
of Artificial Intelligence. On the other
hand, this is not a book one wants to
read from cover to cover. Browsing through
it, reading an abstract here and a paper
there will already allow one to catch
a glimpse of a most fascinating field
of discovery.