Logo IMG
HOME > PAST ISSUE > May-June 2001 > Article Detail


Computing Comes to Life

Brian Hayes

People have been recruiting other species to serve human needs for at least 10,000 years. We have turned plants into crops and animals into beasts of burden; even microorganisms have been pressed into service as fermenters. Yet until now no nonhuman species has ever been harnessed to do intellectual work on our behalf. That could change. Biologists and computer scientists have designed digital logic gates based on the metabolism of living cells, with the aim of eventually building a computer out of colonies of Escherichia coli or some other single-celled organism. But perhaps build is the wrong verb here; the plan is to grow or breed or culture a computer.

The idea of a bacterial computer is not in itself quite so outlandish as it may seem on first acquaintance. In principle, computing machines can be made out of almost anything, from billiard balls to Tinker Toys, and there is no reason that lipid sacs of proteins and nucleic acids should not also qualify as computer building blocks. From the lofty and austere perspective of computer science, an agar plate coated with microscopic bacteria is not much different from a silicon wafer etched with microscopic transistors. If the components can store and manipulate information in a few basic ways, they can compute.

So much for the lofty and austere view of computer science; but there is also computer engineering to be considered, and the questions asked in that discipline are more down to earth. Can living logic gates be strung together in networks large enough to perform an interesting computation? Can they run fast enough to complete a task within a human lifetime? Can they be made reliable enough to produce consistent and correct answers? Can biocomputer engineers cope with all the distinctive failure modes of living organisms—disease, predation, parasitism, senescence, death? (In this context the threat of a computer virus is more than a metaphor!) It's fair to say that practical applications of biological computers are a long way off. And yet skeptics might keep in mind that the historical record of domestications is a vast catalogue of unlikely-seeming successes.

comments powered by Disqus


Subscribe to American Scientist