Von Neuman machines and nanotechnology

[Shangorilla]

I need some help in understanding how self replicating machines make sense. I'm very interested in Dyson's idea of the astrochicken and self replicating space probes along with molecular nanotechnology but I'm still having a hard time grasping how this work. Would a self replicating space probe only make a copy of itself and nothing else or what? And why? Also universal assemblers, how does this work? Like the molecular assemblers in The Diamond Age, how does this work? Is there a computer network that tells all the nanoassemblers what to do or are they themselves intelligent?

 

[cosmictraveler]

http://www.sciam.com/article.cfm?articleID=000BEC4E-39F3-1D6E-90FB809EC5880000

http://www.foresight.org/nanodot/?p=2532

 

[Stryder]

Well you ask the same questions of DNA and it's replication, of course DNA isn't intelligent in the sense of it making decisions but it does follow a predetermined plan in the form of it's RNA blueprinting. Of course for DNA to be manufactured it has to be a component of an overall organism since the intelligent needs of nutrition and it's processing to basic building materials is done at that level.

Self replicating probes/nano's would have to work under a similar method, as a single atom (atom meaning; small composite) of an overall artificial organism they would only contain a very simplified programming which they can not complete alone (A bit like the RNA blueprinting) they would need to work together in neural network clusters to achieve the replication goal as a whole, thereby existing as an artificial organism.

Of course this is just the philosophies that I'd pose.

 

[Shangorilla]

But does a self replicating space probe mean it makes copies of itself? I wanna know what the point of that is.

 

[spidergoat]

It theoretically could. That way you could just send one. It would build another of itself, and set off in another direction. You would have an exponential increase in their numbers. Of course, once they reached a promising destination, they would also have to perform whatever function you wanted them to do then, like construct a communications system to let you know what it found.

 

[weed_eater_guy]

Might also make sense to let them have a means to evolve to better exist in their surroundings. Oh, wait, that would be sorta like creating a life form from scratch, cool...

 

[eburacum45]

A self-replicating probe would need to create some heavy duty machinery in order to build a complete replica of itself and launch it to another star. Once the clone probe is built that machinery is still around- it could be used to build another probe, or to build other infrastructure. One useful thing to build would be a radio antenna so that the probe building machinery could receive more instructions from the home system.

Given new instructions the probe machinery could start building anything.

 

[BenTheMan]

Please forgive a bit of trolling From xkcd.com:

 

[weed_eater_guy]

makes you wonder, maybe we're the evolved result of an alien von newmann machine that landed on this planet long ago in the form of a bacteria? idk, just a neat thought.

 

[weed_eater_guy]

matter of fact, I'll be money someone's written something on that idea, anyone familiar enough to recommend something for me? just seems like a really neat idea

 

[eburacum45]

You might be interested in thi list of 'self-replicating probes in fiction'
http://en.wikipedia.org/wiki/Self-replicating_machine#Self-replicating_machines_in_fiction

 

[eburacum45]

And then there is a couple of fiction pages I have made myself;
http://www.orionsarm.com/topics/Cybyota.html
http://www.orionsarm.com/xenos/Jeepers.html
the Cybiota are a swarm of biological self-replicators sent out long ago by some long forgotten alien civilisation; they have spread across a few planets, but over time have generally forgotten what they were supposed to be doing.