In Survival of the Sickest, Sharon Moalem argues that contemporary human (and other animals') susceptibility to disease typically reflects adaptation strategies of our ancestors to different conditions. It is widely known, for example, that the greater susceptibility of persons of African descent to sickle-cell anemia was an adaptation that enabled their ancestors to survive malaria. The gene for sickle-cell anemia looks like and is a misfortune for people living in societies with means to combat malaria, but it can be a blessing for those who are otherwise vulnerable. Moalem shows how a great many of our genetic predispositions to disease have this feature.
Among Moalem's most interesting observations is a point he makes about pathogens. He explains how it is not generally in the "interest" of a parasitic organism (such as a virus or bacterium) to kill its host. By this measure, the common cold viruses are remarkably successful. By evolving into mere nuisances rather than deadly plagues, they have ensured themselves a plentiful stock of hosts. From this fact, Moalem suggests that much of modern medicine's approach to pathogens may be counter-productive. Antibiotics that aim to kill all of a certain kind of germ end up selecting for the drug-resistant strains, which are often more virulent. We might do better, he suggests, to "encourage" pathogens (such as HIV) to enter into a more symbiotic relationship with us. Moalem then outlines how one could accomplish this task.
And that brings me to computer viruses, or more precisely, worms. With computer security experts still confounded by the conficker worm (as explained here), they may be adopting a strategy that causes its author or authors to become more virulent. We don't exactly know the point of the botnet that conficker uses, but the best guess is that it is commercial: By turning innocent computers into zombies, conficker's master or masters can then sell space on its botnet to spammers and others. The computer security experts are trying to eradicate conficker and to apprehend those behind it. If that succeeds, great, but there is a significant chance that these efforts will fail or, in the process, lead the confickerers to engage in electronic blackmail or terrorism--effectively commanding the botnet to destroy valuable data or disrupt vital programs as retribution.
But what if we learned to live with conficker? With computer memory becoming ever-more plentiful, it might be possible to treat botnet infections as a kind of inevitable nuisance like the common cold. We would simply accept as the cost of doing business, some level of zombification of our computers. We would still treat the symptoms and take precautions against new infections, but by lowering the stakes, we would avoid prompting the confickerers into raising the ante.
I'm not suggesting that this approach would necessarily work. There are some disanalogies between actual germs and computer germs (and worms). But I'd feel better knowing that the security experts had given some thought to the possibility that their all-stick-and-no-carrot approach might be counter-productive. Oh, and by the way, this insight can apply to political leaders as well. Just ask Muammar al-Gaddafi.
Posted by Mike Dorf