Learning from functionalism: prospects for strong artificial life
Elliott R. Sober · Cambridge University Press eBooks · 2010
TWO USES FOR COMPUTERS There are two quite different roles that computers might play in biological theorizing. Mathematical models of biological processes are often analytically intractable. When this is so, computers can be used to get a feel for the model's dynamics. You plug in a variety of initial condition values and allow the rules of transition to apply themselves (often iteratively); then you see what the outputs are. Computers are used here as aids to the theorist. They are like pencil and paper or a sliderule. They help you think. The models being investigated are about life. But there is no need to view the computers that help you investigate these models as alive themselves. Computers can be applied to calculate what will happen when a bridge is stressed, but the computer is not itself a bridge. Population geneticists have used computers in this way since the 1960s. Many participants in the Artificial Life research program are doing the same thing. I see nothing controversial about this use of computers. By their fruits shall ye know them. This part of the AL research program will stand or fall with the interest of the models investigated. When it is obvious beforehand what the model's dynamics will be, the results provided by computer simulation will be somewhat uninteresting. When the model is very unrealistic, computer investigation of its properties may also fail to be interesting.