Extreme Cyber Edge
Genetic Algorithms and the Evolution of Virtual Spaces
Figure 13.7.1: bi7a.jpg
Virtual Spaces Evolved from a Phenotypic Algorithm
As we saw in the previous section, organic artists using generative
methods are able to create wonderfully interesting virtual worlds
and creatures in them. Biota.org's project Nerve Garden illustrated
how to populate a virtual world with simulated plants. Why then
could you not generate a whole virtual world, making three dimensional
structures "on the fly"? Well, you can! One great example
of this is Eric Bucci's work at the University of Texas at Austin.
Working with Professors Marcos Novak and Michael Benedikt, he
has generated a large array of virtual architecture, one example
of which is seen in the figure above. Michael Benedikt's seminal
book Cyberspace, First Steps back in the early 1990s (and
listed in our Bibliography) mentions generative architecture,
notably Marcos Novak's "Liquid Architectures in Cyberspace".
Eric was completing his masters degree and sent us this description
of his work on the "embryology of virtual spaces":
This images is one generation in the evolution via genetic algorithm of a virtual space, starting at zero generations with a population of eight random members, through 60 generations of evolution. Both structure and texture of the model are produced by the genetic algorithm, but only structure is evaluated for fitness.
It is important to note that this algorithm is phenotypic rather
than genotypic. That is, the characteristics which are passed
from generation to generation are contained within each generation's
physical structure rather than within a representative code structure,
like a chromosome set. In effect, it is a genetic algorithm without
genes. The algorithm which produced these structures was coded
in Mathematica, a mathematical software package/programming language.
Eric then goes on to say something about the application of generative
virtual architecture:
The recent emergence of virtual environments as a realm of
architectural design and inquiry precipitates questions regarding
design philosophies in virtual versus physical environments. The
search for valid architectures of Cyberspace will become increasingly
important as the technology which allows its existence becomes
more pervasive.
Concurrently, the continuing development of computer technology
has shed light on the use of algorithms, most recently genetic
algorithms, as an avenue for addressing problems in a wide range
of disciplines, including biology, economics, and reluctantly,
architecture. The use of the genetic paradigm, with its mechanisms
of mutation, fitness evaluation and selection offers a mode of
design with distinct differences from traditional methods. My
thesis proposes that genetic algorithms can provide an enhanced
link between design and construction in architecture, a framework
for generating and managing complexity in design, and an effective
means of access to inhabitation of Cyberspace.
Visit Eric Bucci's home page at: http://mather.ar.utexas.edu/people/alumni/bucci/bucci.html
Digi's Diary Let Life Out
(Or In)!
So, why would life want to squeeze through the cracks into digital
space anyway? Well, for billions of years life has struggled to
build up complex bodies and ecosystems only to have them destroyed
by one mass extinction after another. Whether it is an asteroid,
overheated volcanoes or a too-successful member of Earth's biota
(i.e. us), mass extinctions erase the fruits of immeasurable evolutionary
toil. Some argue that mass extinctions are a kind of spring cleaning
that opens up new opportunities for evolution.
However one looks at it, the inevitable and final mass extinction
is on its way. Far in the future when the sun's hydrogen fuel
is spent and it enters its red giant stage, our puny Earth will
be consumed as an afterthought. So ultimately all this experimentation,
striving, living and dying is going to come to an end, without
much of a trace left behind. This destiny is assured if Earth's
life remains pinned down by gravity onto this one planet. Surprisingly,
the digital realm gives life some properties it could use to escape
this ultimate extinction. For one, life represented purely as
information structures can recombine and reproduce much more rapidly
than its molecule-based forebears. Another point is that digital
life forms would not be tied to the supply of atoms or have to
compete with the body plans and survival strategies that occupy
all the current niches in the physical world.
Another potent property of life in the virtual is that it can
travel through free space or conduits such as wires or optical
fibers at the speed of light. Aim a laser at just the right spot
on the moon and it will bounce off a corner reflector left by
an Apollo crew back in the 1970s. If you packaged up your artificial
life as a datastream you could laser it on a neat round little
trip to and from the lunar surface. Well, you might say, that
hardly counts as sending life to the moon, if there is nothing
to receive and keep it there. Yes, I would agree, but I would
counter that the manned Apollo lunar landings themselves did a
similar kind of bio-reflection. They used enormously costly effort
and large cumbersome vehicles to send men there only to return
them back to the Earth.
Galileo's new brain
The only examples of working off-planet receivers are unmanned
planetary probes like Galileo. Galileo is a large robot spacecraft
now orbiting the Jupiter system. Just after its launch from Earth,
Galileo had an unfortunate accident with its main antenna, which
failed to open. Mission controllers had to decide whether to scrap
the mission or work out a fix. Galileo had one remaining functional
antenna, known as a low gain antenna, through which all communications
with Earth now had to pass at an excruciatingly low bit rate (far
slower than your modem). Mission planners decided to rewrite a
great deal of the software that controlled Galileo's primitive
vision, communications, and other sensing systems. Uploading this
software took days but after it was completed, Galileo had a new
brain. As you read this, Galileo is now carrying out its mission
to peer down at the moons of Jupiter, and study the giant planet
and its electromagnetic field and rings. Galileo sees in a series
of 'jailbars' which allows mission controllers to quickly scan
a scene and tightly focus the camera on features they find interesting.
Galileo also compresses its data much more than its original programs
were designed to do, as this technology improved so much in the
few years since Galileo was launched. In a strange way, a little
piece of the mechanism of Earth's life forms was transmitted out
the Jovian system and stuck.
Of course, the software running inside Galileo will itself be
turned off when the spacecraft reaches its end of life. This software
and its host, Galileo, both lack the ability to reproduce and
carry on. One could imagine some far future date when such vehicles
and the protoplasm of software and data within them have the ability
to use local resources and add on to the vehicle. This type of
craft, which would undoubtedly look radically different that anything
we might conceive of today, might employ nanotechnology to stream
in molecules and weave or extrude new structures.
What would a nano-biota craft make? I think that a kind of artificial
lichens may be a good bet. On Earth, lichens colonize the barest
rock and live by virtue of colonies of sunlight absorbing bacteria
which give them their basic energy source. The lichens also use
acids to break down the rock and prepare the ground for other
life forms to follow. A sort of nano-lichen engineered to live
on the surface of asteroids or comets or any small particle floating
in the Solar System would absorb unending solar radiation and
grow its own RAM. These lichens would bud spores with multi million
year lifespans which can colonize other asteroids or drift into
other star systems.
A little help from our friends
So why is he so down on humans, you might be asking? I think the
human race might find their destiny in space but I don't think
that human beings will sail forth on Star Trek-like ships any
time soon. These ships, with their nineteenth century pipes and
valves will be prone to failure. As can be seen by Apollo 13,
the space shuttle Challenger, space station Mir and any number
of space systems, conditions beyond the biosphere are harsh and
system failures over time are highly likely and often fatal to
a crew of biologics like us. So what do we need to settle the
Solar System? Perhaps a little help from our friends.
Some argue that Earth's atmosphere and oceans were processed by
early single celled life forms which made them habitable by later
more complex life. Could not digital/nano-life forms prepare the
infrastructure that would allow humans to safely and cost effectively
settle other parts of the Solar System? Could these life forms
crack oxygen and other gasses from the lunar regolith, could they
grow energy generation matrices? Perhaps this type of life form
would serve needs inside the Earth's biosphere, helping to clean
up pollution and generate food to allow us to free up agricultural
land. Of course nano-biologics, as you might call them, could
also colonize the our own world to the detriment of us and our
brethren here. Don't panic, though, I think we might have a few
decades to use our computer spaces to evolve some forms in simulation
before we can fabricate them in atoms.
So where is this 'digital life' now?
All of this seems so far off in the future, so what is happening
today? We have all heard of computer software viruses and
the debates over whether or not they constitute 'real life' or
not. Experiments like Tom Ray's Tierra (see: http://www.hip.atr.co.jp/~ray/tierra/tierra.html)
seems like a very convincing life-like ecosystem. A million ten
year olds with genetic hacking tools may ignite a 'digital Cambrian
explosion'. One of the goals of the Biota Nerve Garden project
described earlier in this chapter is allow ten year olds to hack
L-systems and grow their own plants in virtual worlds.
Of course the digital primordial soup is a fragile and ephemeral
space and relatively tiny compared with the Earth's ancient oceans.
But as far as the properties afforded life by digital representation,
human beings, acting as surrogates, may be providing life its
one best chance to break the bonds of Earthly limitations. Perhaps
as we perpetrate our own mass extinction, we can open this new
niche, and let life in (or out, depending on how you look at it)?
What does this have to do with avatars?
After all this you might be asking: so grass will grow in virtual
worlds and force all of us avatars to mow the lawn every week.,
so what's the big deal? Well, life interesting not only because
we are surrounded by people and buildings but by the other living
things. In modern day life it is sometimes easy to forget that
the Earth is inhabited by a rich biota. We will only really miss
it when it is gone, it seems. So digital life in virtual worlds
will make them richer and more interesting places. Maybe face
to face with a strange and evolving biota on our computer screens
will remind us how fascinating and precious life in the real Earth
really is.
So what about the virtual E.T. ?
Some people claim that we won't have to wait for UFOs to land
to meet aliens, we will meet them as a form of digital biota on
our computer screens! Do I think we will soon see the virtual
E.T.? Well, years ago scientists thought that we could create
artificial intelligence through software. It turned out to be
a lot harder than anyone thought, as we had no good working definition
for intelligence or even for thought processes. My own humble
opinion is that we will see some simple undeniably alive in digital
form some time early in the next century. I suspect that it will
look a lot like a virtual slime mold. A slime mold is a
fascinating form of life in that individual cells come together
to form a single colonial organism for a while before reverting
back to individual cells. Slime molds could be a great model for
net-based life forms as the layout of servers and the pipes between
them maps well into an organism made of distributed parts that
come together at points. Search agents traversing the Internet
today and creating central indeces seem to suggest that this form
is possible.
Digital biota should provide us all a great deal of entertainment
and a chance to learn about the rules guiding living things. It
will be a long time before some fully evolved form walks out of
the net and onto the surface of some distant world. If nothing
else, this vision might provide us at least one answer to the
question: why are we doing this?
[begin sidebar]
From the mission statement of Biota.org: Life Crossing a New Barrier
By Bruce Damer, November 29-30, 1996
In its 3.8 billion year existence, the great collective of Earth's
life forms has pushed through barriers to express itself into
new realms. It may have begun with the barrier of cold water around
hot, chemical rich oceanic vents. Another likely barrier was the
radiation-exposed top layers of ocean. The water-air barrier,
the water-land barrier and then the land-air barrier were all
breached in turn, allowing life to flow through. Living things
were carried across each barrier by other living things. Those
surrogate barrier breachers may have been nature's greatest opportunists
or just forced into that role by accident. The first waves of
simple living forms sweeping through each barrier transformed
the new territories and set down the infrastructure for more complex
life to follow.
Few would disagree that the Earth-space barrier is the gateway
to another great territory. Some argue that thoughts themselves
are a viable form of life, expressed across the mind-meme barrier
through language. We posit that another significant barrier is
now being breached. It might be called the DNA-digital barrier
or the atomic-bitomic barrier. It is the expression of life into
the pools and channels of Earth's collective compute spaces. Some
may argue that software viruses, evolving genetic algorithms and
other forms of what we call 'digital biota' are merely simplistic
emulators of real life processes. Others claim that all software
is a bona fide form of life, competing ferociously for system
resources and human attention.
Regardless of whether we are in the digital pre-biotic era or
not, we believe that the crossing of the atomic-bitomic barrier
is inevitable and is a valid expression of life into a new and
viable ecological realm. As human beings perpetrate a mass extinction
on the Earth, so may the collective force of life be pressing
against the bitomic barrier, using us as an unwitting surrogate
to access a rich new realm for biodiversity. The bitomic expression
of life may also be closely linked to the Earth-space and mind-meme
expressions.
Freed from the constraints of chemistry and able to travel literally
at the speed of light, bitomic life could transit the solar system
in hours, but would only take hold if some storage and expressor
mechanism was out there to receive it. The digital realm may itself
be a kind of surrogate, for if molecular nanofabricators one day
permit the re-crossing of the bitomic-atomic barrier, like the
ancestors of whales returning to the ocean, digital biota will
evolve and emerge well adapted to the vast ocean of space. Thus,
life finds a way through the keyhole of human technology escaping
both the bounds of Earth and the mortal coils of the double helix.
If digital biota begin as pure creations of our reason and our art, then they are close cousins to memes, those virulent thoughts which step from mind to mind. Perhaps digital biota will be ideal carriers of thoughts, capturing whole ecologies of memes and allowing us witness and better understand the seething agent societies of our own minds. Our survival depends primarily upon improving this understanding.
[end sidebar]