The power
of computer processing could one day solve the riddle of life's origin.
Scientists
think life appeared about 4 billion years ago, and ancient rocks on Earth can
give us some idea of what the environment was like. Life may
have originated in an ocean rich in chemicals. This primordial soup may
have been simmering, or it may have been zapped by lightning. Certainly energy
of some sort must have helped drive a simple chemical system into a more
complex state. But the clues are few, and the picture remains hazy.
Enter the
Evogrid, a computer creation concept that would be a digital version of the
primordial soup. The EvoGrid was dreamed up by a group of international
advisors and Bruce Damer, the founder of a research company that creates 3-D
spacecraft and mission simulations for NASA and the space community. Damer and
his chief architect, Peter Newman, are developing the EvoGrid concept by
adapting GROMACS, a powerful open source molecular dynamics simulator
originally developed at The University of Groningen in the Netherlands.
Each
virtual particle within the Evogrid's simulated liquid soup will have
particular physical properties, and will behave accordingly.
"We will be
constructing a model of a 'toy universe', which has approximate properties of
the early oceans on Earth," says Damer.
Cooking
creation
With a
laundry list of basic physical properties entered into the starting parameters,
the simulation would allow artificial nature to take its course. Interactions
and connections between particles should occur, and ever higher levels of complexity
may arise from the most basic elements.
Much like
SETI@home's screen saver, which enables computers at home to search for signals
of extraterrestrial life within volumes of astrophysical data, the Evogrid
is conceived to have volunteer computers become part of an interconnected grid
for maximum processing capacity. Damer hopes to eventually get a million
computers hooked into the grid.
These
computers would receive data from the EvoGrid simulation engine. The simulation
would essentially consist of a vast virtual ocean of interacting numbers that
would model the time before complex life forms emerged. To know whether
self-organization is occurring, the program would look for persistent patterns
within the data.
"If a
vesicle, or a ball of particles has formed, you would be able to detect that,"
says Damer. "If a string of particles began to replicate, that would be easy to
track, or if particles began to combine in a long chain of reactions, that
would be important but tougher to recognize."
It is
thought that some combination of a lipid container (vesicle), strings of
molecules (genomes) and metabolic reactions led to the development of life. The
Evogrid won't produce visual images of the combined effort of all the linked
computers, because it would slow the processing down too much. However, home
users would see a visualization of what is being observed in their own small
patch of the EvoGrid.
Damer notes
that present-day computer simulations run much more slowly than chemical
reactions, but he anticipates that in the next 20 to 40 years, with the help of
millions of microprocessors, an entire cell could be simulated in cyberspace.
"Nils
Baricelli wrote an artificial life program for the first modern computer in
1953, and to some extent we haven't gone much further than his original
experiment," says Damer. "We shall see how far the EvoGrid can go, using
millions of the descendents of the original Von Neumann
machine."
Tinkering
with life
Damer
envisions two possible versions of Evogrid: a hands-off "Origins" version, and
an experimental "Intelligent Designer" edition that would allow people to
tinker with the simulation. Damer says the ID edition of Evogrid could include
a "miracle module" that would allow users to play God in their attempts to
create proto-life. The Origins edition would be the focus of the science,
however, with strict controls to shield the experiment from any guiding human
influence.
Damer muses
that "in its ultimate incarnation, a much more powerful EvoGrid would allow us
to pose the question: where in this universe or others might life exist and at
what level of complexity?" Damer thinks an EvoGrid tuned for SETI and
astrobiology could be used to simulate extraterrestrial
environments and address the question of whether life could have emerged
there.
Even if the
EvoGrid managed to generate some virtual but convincing life forms, either
through random or directed means, "the numbers will always be numbers," says
Damer. "As Baricelli wrote over fifty years ago, they will never be living
organisms."
But Damer
dreams big, and he thinks someday the creatures generated by the Evogrid could
be re-created chemically. A virtual scanner could be devised to break down the
computer-based creature into its digital body parts, and then that information
could be used to try to build the same creature out of real chemistry in the
lab. Of course, this step of the experiment would rely on technology that does
not currently exist. "Life is more than the sum of its parts, and you can't
just throw the necessary chemicals together and expect a life form to emerge,"
Damer says.
However,
researchers are hard at work trying to recreate all the biochemical steps
necessary to synthesize a kind of proto-life in the lab, so perhaps this
possibility is not too far over the horizon.
Looking
even farther into the future, Damer thinks that far more advanced EvoGrids,
paired with "ChemoGrids", could be used to create a new genesis of cyber-physical
life forms to colonize asteroids, or to terraform Mars into a more habitable
planet for humans.
He expects
that other intelligent civilizations in the universe probably harness the power
of evolution to solve difficult problems such as creating habitable zones to
colonize. "This is a common theme in science fiction, but science fiction tells
us what could be possible someday," he adds. "The way for us to get there is to
start with simulation and ride the wave of ever greater computing power."
advertisement
