boston.com News your connection to The Boston Globe
WHAT WOULD DARWIN SAY? | LEE SPECTOR

And now, digital evolution

RECENT developments in computer science provide new perspective on ''intelligent design," the view that life's complexity could only have arisen through the hand of an intelligent designer. These developments show that complex and useful designs can indeed emerge from random Darwinian processes.

Most scientists agree that the argument for intelligent design, which is over 200 years old, was put to rest by Darwin in 1859 and by the subsequent triumphs of 20th century biology. Nonetheless, President Bush recently advocated teaching intelligent design in schools along with evolutionary theory. School boards around the country, most notably in Kansas, are staking out similar positions. What can account for the persistence of this long-discredited idea?

Scientific illiteracy is certainly part of the explanation, but other factors are also at play. Prominent among them is the fact that discussions about evolution are usually discussions about the origins of the discussants themselves. We tend to hold our own species in high regard and to look down on random and mechanical processes. Our appreciation for the magnificence of humanity, and of all life, is well grounded but it may also blind us in ways that we are not blinded when studying rocks or electricity.

Thanks to technology, however, we can now explore evolution without discussing ourselves and without even discussing life. We can do this by building evolutionary processes into computer programs. When we do this we do not find that our appreciation for the magnificence of life is in any way diminished. Rather, we find that our appreciation for the power of evolution is amplified.

A growing sub-field of computer science is devoted to ''evolutionary computation." The user of such a system specifies the ingredients that can be used and how the ''goodness" of any particular design can be measured. The system then creates and tests thousands or millions of random combinations of the ingredients. The better combinations are allowed to produce ''children" by mutation (random changes) and recombination (random part-swapping). This often produces, after many generations, genuinely novel and useful designs and inventions.

Evolutionary computation has proven to be useful for solving practical problems. It has been adopted by researchers and engineers, and it is the focus of scholarly journals and international conferences.

One of the major conferences offers cash prizes for ''human-competitive" results; that is, for cases in which evolution has out-designed human beings. To win the prize contestants must show that their systems equaled or improved upon the performance of humans as measured by such criteria as patents and peer-reviewed publications. In 2004 I shared the ''gold medal" with a team from NASA that evolved an antenna for the Space Technology 5 mission.

My entry involved the evolution of quantum computing circuits, which are difficult for humans to understand or design. More to the point, they are extremely difficult for me to understand or design, and I could never have produced the results on my own. I am not a designer equal to that task, but evolution is. I created the ''primordial ooze" out of which quantum circuits could grow, and I wrote the programs for random variation and selection. But evolution did the heavy lifting.

Of course biological evolution and evolutionary computation differ. Engineers using evolutionary computation specify explicit measures of ''goodness" that govern selection, while biological selection is governed only by survival and the ability to reproduce. In addition, biologists now understand that processes other than natural selection, for example symbiosis and influences during development, also contribute to evolution. But evolutionary computation and biological evolution are both fundamentally driven by random variation and selection, and the successes of one hint at the power of the other.

It is easy to appreciate the power of selection operating on random variation when it is stripped of its emotion-laden connections to human origins and is shown to be capable of designing complex solutions to difficult problems. If one extrapolates this power to a system the size of the Earth, then it may not be such a stretch to imagine that evolution could produce the stunning complexity and beauty of our biosphere. Viewed in this light, Darwinian evolution is itself a designer worthy of significant respect, if not religious devotion.

Lee Spector, professor of computer science at Hampshire College, is the author of ''Automatic Quantum Computer Programming: A Genetic Programming Approach."

SEARCH THE ARCHIVES
 
Today (free)
Yesterday (free)
Past 30 days
Last 12 months
 Advanced search