Endless Factories

Over the past few weeks, I've been covering lots of recent developments in the worlds of nanotech and 3D printing, as well as a brief overview of the concept of fractals (as it pertains to self-replicating patterns). Today, however, I'd like to briefly address an intriguing issue in the world of replication...that of "self replication".

Previously, I mentioned a little bit about the RepRap machine, which can potentially reproduce parts for yet another RepRap machine. The end goal, of course, is to create a machine that can fully replicate itself. So far, the videos I've seen only show the machines making parts, but not actually assembling them. Now, the concept of self-replication is just that...a machine reproducing copies of itself and even assembling that copy without any human intervention. The copying process would likely be driven by software code...much like DNA is used by a cell as a set of instructions.

This idea is all fine in theory, but it's doubtful there will be any error-checking in the copying process. A self-copy is only going to be as good as the original master machine itself...which means if there are any hidden "bugs" in the original, they'll be in the copy, too.

In addition to large scale attempts at self-replication, the race is also on to create self-replicating nanobots. Assembly is much more difficult at the nano scale, and although significant strides have been made to create nanoscale machines, self replication presents a new set of challenges. Despite the impressive engineering involved in such a creation, it seems that the tradeoff for size will also result in an increase in vulnerability and fragility, at least at an individual level.

How does one overcome such an obstacle? One way would be an increase in the volume of individual machines. This idea can be easily found in the world of biology, but in terms of nanotechnology, the best analogy might be bacteria. On the surface, this sounds like a good idea, but there is the potential of a "gray goo" scenario unfolding (as described by Eric Drexler). This scenario involves self-replication that gets completely out of hand...to the point that it could cause chaos on an enormous scale especially if the devices consume local resources in order to reproduce themselves.

Science fiction has a colorful history when it comes to describing possible scenarios with self-replicating machines...from Stargate SG-1 to Star Trek to books/short stories such as Philip K. Dick's Autofac. This is one topic I hope to explore in the next project I'm working on, Fractal Standard Time.

Unfortunately, I think research into this area will end up being more of an afterthought...well after the actual devices are created and released into the world. It would be one thing to deal with RepRap machines run amok, but nanomachines? Depending on the creation, the consequences of flawed design in both hardware and software would be unpredictable at best and maybe even unstoppable at worst. How can you prepare to fend off a device you can't even see?

Next time I'll discuss one of the potential hidden flaws to these theoretical machines: software bugs.