Scientists are using computers and robots to mimic the behavior of living organisms. In the process they claim to be discovering the processes that gave rise to life itself and drove evolution. They might be right.
Technology journalist Ward (Computer Weekly) tells of human efforts to create artificial life. Genetic engineering is not involved, and the end result is not a living creature but a computer program, less often a robot. In this ingenious game, “life” (a collection of dots on a computer screen) follows simple rules of life and death: as time passes, the dots grow, move, oscillate, and change shape in unpredictable ways. It may not be life, but it’s creepy. A more sophisticated program (“Tierra”) exists only inside a computer memory. Several Tierra programs compete for computer resources, and, sure enough, they evolve. Programs thrive that grow more efficient and multiply; those that lag behind die out. Smaller programs are more nimble, so they increase, but below a certain size the program can’t reproduce. Mutants appear below this critical size but survive by using the reproductive commands of other programs. Are they parasites? Critics insist that no computer can be “alive” or “intelligent.” They have an easy job because they deal with progress by moving back the goalposts. Thus, in the past critics pointed to chess as an activity that requires the mysterious complexities of a living brain. Now that computers easily beat grandmasters, the chess argument has been dropped.
The best work on this subject is Steven Levy’s Artificial Life (1992), although it’s getting long in the tooth. Ward covers the same ground and features the same colorful characters, but lacks pictures and diagrams to illustrate difficult concepts. Read Levy first.