Life preserves informationas defining both ``life'' and ``information'' in terms of each other. It is certainly plausible to claim that ``All living systems do an effective job preserving information''. Is it completely absurd to say that ``All effective information-preserving systems are living systems''? We ground out the mutual recursion in cases where we have other prima facie reasons to describe a system as ``living'' or as ``preserving information''---and our understanding of the dual role of DNA, as both an active catalytic controller of chemistry during the existence of a cell and as a passive reaction product during cell copying, provides one such base case.
Such a high-handed approach, while riding roughshod over all sorts of important issues, offers a way to unite ``life is selfish gene-copying'' advocates (e.g., Dawkins, 1989) with ``life is self-production'' advocates (e.g., Maturana & Varela, 1980), viewing copying and maintenance as the two fundamental strategies for preserving information. On some of the standard challenges for definitions of life, the view would include mules, exclude fire, and probably leave crystals on the margin, depending on how much ``information-theoretic'' information we expect to find in a crystal---which won't be much, if the crystal is pure. Impure or semi-crystalline materials, such as, say, integrated circuits, are of course another matter.
On the prospects for life in manufactured computers, the view is that, rather than being an esoteric research topic, that is a prosaic, long-established fact. Popular software programs today are preserving their information spectacularly, with population sizes in the millions and booming; malicious computer viruses are harder to measure though detections of new strains are booming as well (see, e.g., Kephart, 1994). The emergence first of affordable personal computers and now of mass-market computer networking adds urgency to the real research question: With the great flexibility of programmable computers laying before us, what kind of artificial life do we want?