VIRTUAL ORGANISMS: The Startling World of Artificial Life

VIRTUAL ORGANISMS: The Startling World of Artificial Life MARK WARD. St. Martin's/Dunne, $23.95 (320p) ISBN 0-312-26691-X
Artificial intelligence research has tried to make machines that think; the newer and in many ways more exciting field of artificial life ("ALife") seeks computers and computer-riv en machines that work like-or arguably in some sense are-living things. ALife "encompasses software simulations, robotics, protein electronics and even attempts to re-create the world's first living organisms." This compelling and easy-to-follow volume from the Daily Telegraph (U.K.) tech journalist Ward picks up where Steven Levy's Anti facial Life (1992) left off, surveying recent and classic AIr ife work in all its subfields. Bell Labs researcher Andrew Pargellis's "computer simulation of a primordial soup" produces "working, replicating programs" analogous to the self replicating molecules that colonized the early Earth. John Horton Conway's computerized "Game of Life" produces "CellularAutomata," self perpetuating, evolving patterns that model biological evolution. Cambridge scientist William Walter's 1950s robots "Elmer" and "Elsie," he claims, chased each other like cats and learned tricks like dogs: inspired by them, MIs Rodney Brooks makes robots that can explore the real world, "solving the same problems that animals face." Programs that replicate, mix with other programs and generate somewhat different successors mimic the sexual reproduction that has made possible much of our evolution: these programs, called "agents," may someday run telephone networks and other large electronic systems-with catastrophic consequences if they evolve in ways that are bad for us. Though he includes some scary scenarios, Ward is largely upbeat about the scientific and practical future of ALife in all its manifestations. After his sometimes exciting, always accessible exposition, his satisfied readers may learn to love it, too. (Nov.)

Virtual Socioblogy

Virtual Sociology; Artificial life forms inhabiting virtual worlds are nothing new to fans of computer games like The Sims, but rewiring artificial life for scientific research is a new frontier.

Karen Jones. PC Magazine. New York: Sep 20, 2005.Vol.24, Iss. 16; pg. 21

Artificial life forms inhabiting virtual worlds are nothing new to fans of computer games like The Sims, but rewiring artificial life for scientific research is a new frontier. The New and Emergent Worlds Through Individual, Evolutionary and Social Learning (NEW TIES) project is developing a computer-simulated society, observed by experts, which may eventually benefit information technologies, computing systems, AI, and linguistics.

Backed by a group of European universities, NEW TIES will place 1,000 agents in a simulated world on a 50-node network, with the goal of seeing them evolve a culture. "The project can help us find out about the factors necessary for certain attributes of society to come into being," says Professor Ben Paechter of Scotland's Napier University.

Paechter adds that agents will start with few skills and will need to learn how to acquire food and to communicate. Their language will evolve from scratch, says Dr. Paul Vogt of Tilburg University in the Netherlands.

As to how NEW TIES might benefit real societies, Paechter posits that it makes otherwise impossible experiments possible. For example, the agents could be stripped of key resources in their environment, and the effects of the hardship studied in ways that would be inappropriate in a real society. You can track the virtual culture at www.new-ties.org .