I have read that computer programs have been developed which demonstrate that the random process of natural selection can be creative.
Dr. Coss: First of all, evolution is unsupervised, impersonal, unpredictable, and a natural process. That is, nature is all there is. Now, this quote by the evolutionist, Monod, sums it up very nicely: ". . . chance alone is at the source of every innovation, of all creation in the biosphere. Pure chance, absolutely free but blind, at the very root of the stupendous edifice of evolution." Now, if we look at this philosophically, we have to say that natural selection cannot be directed. That is, natural selection is a random process. Now, how would that compare to a computer program? Some of you have seen in the book, Climbing Mount Improbable, a nice display: very complex spider webs, and these spider webs were produced by a computer program. Well, they may do something like producing a meaningful sentence by using a random process to generate letters that will eventually, through just a few generations, produce a meaningful sentence. Now, is that really comparable to natural selection? Remember that natural selection has no goal or purpose. Remember we said that evolution was a natural process without any intelligent force, without any goal in mind. If you look very carefully, you'll realize that any computer program has a very definite target. They know what they want to produce. They know what words they would like to get. They know what shape they would like to have for, say, a spider web. Consequently, since they have a goal in mind, they will conserve anything that is directed toward the goal. For example, they would keep part of a word that is still, of course, nonsense. But they would conserve that if they had the right letter in the right position. Now, you'll see that there is a very serious problem here when we relate this to natural selection, that is, that non-functional intermediates would not be conserved. In the real world, as Dr. Lovegrove was just mentioning, a partial product, something that was not yet functional, would have no selective advantage. And that would, of course, not be conserved at all. So be very careful when you see someone trying to suggest to you that any computer program is a realistic simulation of natural selection. These two are very, very different processes.