Is there life in your computer?

Originally published 5 October 1992

This is the Nineties, right?

Not only must we pro­tect our­selves from germs. We must also guard our computers.

When­ev­er I pop a flop­py disk into the dri­ve of my Mac­in­tosh, the machine is poten­tial­ly exposed to virus­es — dig­i­tal microbes con­trived by unscrupu­lous hack­ers to turn my com­put­er’s mem­o­ry into mush.

A com­put­er virus is a short pro­gram that pass­es from machine to machine, con­cealed in streams of appar­ent­ly inno­cent data. Once lodged in the mem­o­ry of a host com­put­er, the virus pro­gram copies itself, some­times in mem­o­ry-clog­ging num­bers, then looks for oth­er machines to infect. In all of this, a com­put­er virus is uncan­ni­ly sim­i­lar to a bio­log­i­cal virus.

My Mac is not with­out pro­tec­tion. I’ve invest­ed in a virus detec­tion pro­gram that scans every disk that goes into the dri­ve for com­mon virus­es. The pro­gram is a sort of all-pur­pose antibi­ot­ic that rec­og­nizes and destroys alien virus­es before they cor­rupt my data.

How­ev­er, my com­put­er has a built-in modem that allows it to exchange data with oth­er com­put­ers over tele­phone lines. Tele­phone com­mu­ni­ca­tion is virus-safe for humans, but not for com­put­ers. My virus-pro­tec­tion pro­gram does­n’t check the modem con­nec­tion. If the machine at the oth­er end of the line is infect­ed, my com­put­er is at risk.

So far, com­put­er virus­es pose only a mod­est risk to the own­er of a per­son­al com­put­er. Pro­vid­ed we use pro­tec­tion, and exchange data only with com­put­ers we have rea­son to believe are virus-free, we have lit­tle to wor­ry about. How­ev­er, two arti­cles in a [1992] issue of Sci­ence sug­gest a more wor­ri­some future.

Accord­ing to the first arti­cle, many bac­te­ria and virus­es that cause com­mon dis­eases, from pneu­mo­nia to tuber­cu­lo­sis, are devel­op­ing resis­tance to the drugs that have kept those dis­eases under con­trol. Like all liv­ing things, microbes evolve. If genet­ic muta­tions give a microbe some mea­sure of pro­tec­tion against its ene­mies (antibi­otics), the new genes will be favored by nat­ur­al selec­tion. Microbes repro­duce with aston­ish­ing speed — a gen­er­a­tion every 20 min­utes for the typ­i­cal bac­teri­um. Bac­te­ria also have a way of exchang­ing genes direct­ly among them­selves. For bac­te­ria, the equiv­a­lent of a mil­lion years of human evo­lu­tion can be com­pressed into a few months.

Med­ical researchers are strug­gling to devel­op new antibi­otics faster than microbes evolve defens­es. “There’s no mag­ic bul­let,” says George Jaco­by, a spe­cial­ist in infec­tious dis­eases at Mass­a­chu­setts Gen­er­al Hos­pi­tal. “Bugs are always fig­ur­ing out ways to get around the antibi­otics we throw at them. They adapt, and they come roar­ing back.”

The sec­ond arti­cle in Sci­ence report­ed on the third work­shop on Arti­fi­cial Life (ALife), held in San­ta Fe this summer.

The most excit­ing dis­cus­sions at the work­shop con­cerned the pos­si­bil­i­ty of cre­at­ing “real” life in a com­put­er, and whether dig­i­tal life might already exist in the form of com­put­er viruses.

Accord­ing to ALife afi­ciona­dos, life is not just a col­lec­tion of mol­e­cules or cells; more fun­da­men­tal­ly, it is also a set of process­es, such as metab­o­lism, repro­duc­tion, or pro­tein for­ma­tion. These process­es can be sim­u­lat­ed with dra­mat­ic fideli­ty in the mem­o­ry of a com­put­er. Dig­i­tal “plants,” for exam­ple, have been made to grow, blos­som, com­pete and evolve. When the results are dis­played on the screen of a video mon­i­tor, they are dis­con­cert­ing­ly lifelike.

At the San­ta Fe work­shop, Thomas Ray, an evo­lu­tion­ary biol­o­gist at the Uni­ver­si­ty of Delaware, demon­strat­ed a real­is­tic ALife pro­gram called Tier­ra (Span­ish for Earth). A Tier­ran species of life is defined by a sequence of com­put­er instruc­tions that tell the “organ­ism” how to repro­duce and inter­act with its envi­ron­ment, much as DNA seg­ments do for real-life organ­isms. Muta­tions — a change of a bit from one to zero or vice ver­sa — yield new instruc­tions, some ben­e­fi­cial, some fatal. When Ray’s dig­i­tal organ­isms are set loose in the mem­o­ry of a com­put­er, they com­pete, evolve defens­es and yield prog­e­ny that were not envis­aged by the programmer.

Com­put­er microbes copy them­selves at speeds that make bac­te­r­i­al repro­duc­tion look pos­i­tive­ly poky. When dig­i­tal virus­es can spon­ta­neous­ly mutate, com­pete against their preda­tors (anti-viral pro­grams), and evolve new and more vir­u­lent strains, soci­ety is in big trou­ble. If drug researchers can’t keep up with the evo­lu­tion of bio­log­i­cal microbes, what hope will we have of keep­ing ahead of fast-chang­ing com­put­er viruses?

Increas­ing­ly, human civ­i­liza­tion depends on mas­sive­ly inter­con­nect­ed com­put­er sys­tems. A web of elec­tron­ic data wraps the plan­et, con­trol­ling or reg­u­lat­ing many aspects of our lives: Call it the data­s­phere, by anal­o­gy with the bios­phere. In that rich dig­i­tal envi­ron­ment, future ALife microbes will breed and thrive, pre­sent­ing soci­ety with prob­lems as intractable as the scourge of AIDS. A whol­ly new pro­fes­sion — dig­i­tal med­i­cine — will come into being to keep our machines healthy.

Life in the bios­phere and life in the data­s­phere can be remark­ably sim­i­lar. Evo­lu­tion­ary biol­o­gists are learn­ing about real life by watch­ing ALife organ­isms evolve on com­put­ers. And the unscrupu­lous hack­ers who cre­ate com­put­er virus­es will be look­ing to real-life microbes for ever more devi­ous ways to evade our anti-viral defenses.

---

The source code for the orig­i­nal Tier­ra arti­fi­cial life soft­ware is avail­able in the data­s­phere, but please, watch out for virus­es. ‑Ed.