Molecular manufacturing designs have traditionally looked a lot more like machines than like life's biochemistry, and molecular manufacturing has even been criticized for not taking more inspiration from life's techniques. But there is a fundamental connection between them. The digital nature of chemistry (to be specific, molecule-forming chemistry) is what will allow molecular manufacturing systems to build duplicate systems, driving down the cost of manufacturing and enabling high throughput and large products. It's easy to think of life as squishy and analog, or complex and chaotic, but at its core, life is digital too. And it may even be the case that this is what allowed life to form in the first place.
I was recently discussing the origin of life with some biologists and other scientists. During the course of the discussion, I suggested that life might have developed from a soup of simple organic molecules when a fairly simple self-replicating molecule was templated by clay particles. (This conjecture is not original to me, but I think it sounds at least somewhat plausible.)
Someone objected that, during the time needed for the next advance to develop, any tiny perturbation would destroy the pre-life structure. Someone else pointed me at the work of L. L. Whyte[PDF]. Now, Whyte was writing before much was known about molecular biology, and his theories are about as accurate as the early descriptions of electricity as a fluid. But his observations and logic are first-rate, and one of his observations is that there must be some core that implements heredity and provides an organizing principle to organisms, and that core must be replicated with unusually high fidelity.
As a computer scientist who's studied molecular manufacturing for two
decades, I had gotten used to thinking of self-replication as digital
and highly accurate, in the physical domain as well as the computer
domain. Reading Whyte got me thinking, and I realized that indeed, life does depend on near-perfect duplication of information - and it's not guaranteed that any particular process of copying will provide that.
Of course, this is what DNA does. Using the highly nonlinear forces of chemistry, DNA can be copied with an extremely low error rate - orders of magnitude lower than most organic chemistry processes. If this were not the case, then life would devolve - errors would accumulate faster than they could be selected out. But with the ability to produce essentially perfect copies of DNA, the total number of accurate copies can increase with each generation. Then, the small percentage of random mutations that are beneficial will lead to an increase in the total number of improved copies in each generation - at which point the improved copies will out-compete the degenerated copies, and the species will improve.
In discussing the conjecture that life originated with simple molecular self-replicators, many people will not make explicit their assumptions about whether the copying process is digital and accurate, or analog and lossy. It seems clear to me, now, that 1) the copying process must have been digital, because lossy processes could not have produced a lineage of ever-improving self-replicators leading to life; 2) Since the formation of molecules is, in general, a digital process, the theory survives this restriction.
I could say that this means molecular manufacturing is biomimetic after all. But I suspect that the use of digital copying processes in molecular manufacturing owes its inspiration at least as much to computer science as to biology. The bottom line is that digital copying works, and if molecule-forming chemistry were not digital, there's a good chance we wouldn't be here to talk about it.
Chris Phoenix
]]>
