Star Trek: Life in the Machines?

Consider a class of robot with three components: base, arm, and controller. Suppose that each of these components comes in two types, red/green and blue/yellow, and that two robots are fundamentally the same if each of their components has the same color, or each has the complementary color. Thus, a red-based, red-armed, blue-controlled robot would be 'the same' as a red-based, red-armed, blue-controllered robot, as well as 'the same' as a green-based, green-armed, yellow-controllered robot, but not, for instance, a green-green-blue robot.

Suppose some goofball decided to plunk down one of these RRB robots in a room full of robot components (of both colors), with instructions to make more RRB robots. Now, this instance of RRB robots making RRB robots might be termed self-replication, but... really... you're just sticking three things together. There's no effort to gather resources, they're supplied externally. It doesn't count.

Or does it?

Take a look at our planet's foremost self-replicating system: strings of nucleic acids which differ by having a (G)uanine, (A)denine, (C)ytosine, or (T)hymine nucleotide on them. A GATTACA string is essentially the same as another GATTACA string, as well as a CTAATGT string, but not a GTAATCT string. These little guys do not go out in search of five parts Carbon, five parts Hydrogen, five parts Nitrogen, and one part Oxygen, they do not hand-assemble them into Guanine nucleotides, they do not actively give rise to more of themselves. When tossed into a soup with the appropriate enzymes and building blocks, however, it's hard for chemistry to resist the temptation to copy the things, so it doesn't resist. Does DNA fail to self-replicate simply because it doesn't do its own resource gathering or part assembly... because it's not even as active as a robot arm?

Well, DNA does, sort of, do its own resource gathering. In addition to providing a template that's easy to copy in the right circumstances, it also provides information which (again, in the right circumstances) produces organisms which do go out and gather the necessary resources (often by killing other organisms and taking the resources they'd amassed). It also provides information that, in the right circumstances, can be used to set up "the right circumstances". In addition to being easy to copy, DNA has information which (in the right setting) can be used to put together a setting which will both copy the DNA, and be the right setting to use its information.

Ever worked in a place with a photocopier? Ever been passed one of those office jokes that is funnier if you make copies of it and pass it to other people... especially people who work in places with photocopiers? Especially people who also think the joke is funnier if they make copies of it and pass it on? In short, have you ever had in your hands something easy to copy which had information which, in the right hands, made it more likely to both get copied and to have those copies end up in the right hands?

It makes me wonder how the United Federation of Planets is supposed to survive as depicted, given they have 'replicator technology'. It's the office photocopier, with a vengeance. Given a device that can replicate things with molecular accuracy, what are the chances that nothing will ever develop an appeal... that there'll never be a Plush-Tribble replicating fad? What are the chances that nobody will ever replicate something that the right kind of people will want to replicate and pass on to the right kind of people? And, if that happened, what are the chances that the things that were the most attractive to replicate wouldn't get replicated a lot?

Of course, there would be energy constraints. You can't replicate as much as you want for as long as you want, so eventually the constraints might be bumped into, and you could only replicate so many Plush-Tribbles, R/C Model Starships, and Novelty Nose Insignia. What are the chances that, once the limit was reached, the demand for these things would be equal... that
you wouldn't start getting more Starships being replicated than Insignia, for example... and that once more people were participating in (and helping spread) the Starship replicating craze, that the Insignia craze wouldn't get swept away for eventual lack of support? What are the chances that the Insignia craze wouldn't either go away, or change in a way that got people to start replicating Insignias more than Starship models?

Plush-Tribbles, Model Starships... there's no need to claim these things are alive. There's no need to even claim they replicate themselves, though they replicate themselves as much as DNA self-replicates. Not alive and potentially not-replicating though they may be, they are in competition with each other. Changes, even small ones, in the nature of any of them will alter which are dominant and, possibly, which continue to even exist. Not alive and potentially not-replicating though they may be, one should expect to find that, as time goes on, replicator-crazes will become more and more effective at getting people to A> replicate them and B> pass them to other people who will do parts A and B because, if nothing else, as time goes on any craze that doesn't do those things won't stick around, while any craze that does will grow. And whichever crazes are best at achieving those end effects are the ones you should expect to see clogging up replicator resources.

Perhaps a craze becomes easier to replicate by requiring less replicator time/power. Perhaps by being cuter. Perhaps by having internal motors and controls that move the item closer to replicators. Then again, moving into the replicator would do even better. Then again, having a protrusion that hit the 'replicate' button would do even better than just getting in. If for some reason the replication process wasn't always exact (Like that episode where the coffee got replicated before the coffee cup got replicated... oops... sploosh!) then any minor difference that got something more effectively replicated (be it through direct action, or its incidental effects on human nervous systems (i.e. "OOOOh so CUTE!")) would end up becoming more and more common, and establishing a new baseline that new (or just modified) crazes would have to surpass somehow.

Superficially, at least, it would seem that it's not necessary to be alive, or even to take upon yourself the burden of replicating, to be able to evolve. All you need is a system for performing usually-accurate replication and an environment which applies that system non-uniformly... even a little amount of preference would be enough. Looking at Star Trek replicator technology, I wonder why the Star Trek universe isn't plagued by evolving replicator-fads.