In 1989 Tom Ray, an evolutionary biologist with an interest
in computer science, decided to try and combine both disciplines to create
a new world, that would allow him to observe evolution.
Because evolution is all around us, but works at a very slow pace. The
average human life span is as nothing compared to the hundreds of millions
of years nature needed to get from the first primitive life form to the
many complex creatures currently living around us. This makes the actual
observation of evolution rather difficult.
The idea was to write a small self-replicating program, similar in principle to the first living creature on Earth. This program would then be allowed to run. Every time the program would reproduce, there would be a small probability for a mutation to occur, producing a change in the replicated creature. This is the way life in all its complexity developed on Earth. It was hoped that the same complexity would develop relatively quickly inside a computer, at the fast pace allowed by modern micro-processors.
This was in fact an evolutionary approach to artificial
life, which was at that time emerging as a new discipline. Tom Ray contacted
Chris Langton, and was invited to visit the Artificial Life group at Los
Alamos National Laboratories in October 1989. He discussed his ideas with
the different members of the group.
Most of them were skeptical, as they objected that mutation was not a good
thing to use on standard computer languages. These are too fragile, too
"brittle" for such a treatment. Mutation done on computer code
would be almost sure to produce junk. It would be next to impossible for
the first self-replicating creature to mutate into anything but an incorrect
program.
Tom Ray took the argument to heart, but nonetheless went on to try and
implement his idea. The result was, of course, Tierra.