PDA

View Full Version : How far away are we from true self-replicating technology?



KABOOM
2014-Aug-07, 05:26 PM
Much is often said about machine-intelligence exploring the Solar System. Rebuilding itself in "high resource" neighborhoods such as Oort Clouds or other cosmic zones.

Let me frame the question more specifically. We will make the climate much more hospitable than space, right here on Earth. Say we were to erect some sort of automated, robotic devise that was capable of "observing", "recording" and "transmitting" data to a remote command base (several thousand miles away) and we placed such devise in the middle of the desert in Arizona. We could surround it with whatever other form of remote technology that we have at present. For how long could such a devise last and perform to its full functionality without any human intervention?

My uninformed "guess" would be well less than 100 years but welcome more informed opinions.

cjameshuff
2014-Aug-07, 06:51 PM
We arguably already have replication: we've automated the production of the major elements of modern industry. It's just a matter of scaling it down so we can fit all the functionality in a single package...and equipping it with the capability to locate and collect the required resources.

Arizona may not be as hospitable as you assume. The sun's out of sight or at a low angle to the ground most of the time, and there's a dense oxidizing atmosphere: hot and cold things will have to be well insulated, things will need protection from oxidation, some processes will require vacuum. The readily available materials may be less than ideal...there's little in the way of metallic nickel-iron debris scattered around, for example. The gravity, highest of any solid body in the solar system, means it'll take a lot of energy to move material around...

There's some interesting manufacturing technologies that are particularly well suited for space which wouldn't work well in Arizona. There's a variation of the common nickel refining process that has been adapted to 3D printing: nickel reversibly reacts with carbon monoxide to form a volatile compound, tetracarbonylnickel. You could produce nickel objects from scrap and nickel-rich asteroid material, purifying it as you do so. But there's not a lot of nickel-iron sitting on the ground in Arizona, despite the presence of one particularly impressive hole in the ground produced by a metallic impactor.

Hlafordlaes
2014-Aug-07, 07:19 PM
.... There's some interesting manufacturing technologies that are particularly well suited for space which wouldn't work well in Arizona. There's a variation of the common nickel refining process that has been adapted to 3D printing: nickel reversibly reacts with carbon monoxide to form a volatile compound, tetracarbonylnickel. You could produce nickel objects from scrap and nickel-rich asteroid material, purifying it as you do so.

Fascinating. Is it poisonous?

cjameshuff
2014-Aug-07, 08:43 PM
Fascinating. Is it poisonous?

It seems to be one of the most toxic substances used in industry, in fact.

http://en.wikipedia.org/wiki/Nickel_tetracarbonyl#Toxicology_and_safety_conside rations

Noclevername
2014-Aug-09, 05:05 PM
Let me frame the question more specifically. We will make the climate much more hospitable than space, right here on Earth. Say we were to erect some sort of automated, robotic devise that was capable of "observing", "recording" and "transmitting" data to a remote command base (several thousand miles away) and we placed such devise in the middle of the desert in Arizona. We could surround it with whatever other form of remote technology that we have at present. For how long could such a devise last and perform to its full functionality without any human intervention?

My uninformed "guess" would be well less than 100 years but welcome more informed opinions.

Even the best conditions on Earth still have oxygen and occasional water present, and deserts have dust and sand. Those eventually erode and corrode any functional technology.

Still, if price were no object I think a 100 year transmitter could probably be built. But I would not call it today's technology, as significant R&D would probably be required. Also a lot of engineering. In some ways, I think the outermost Solar System, with its vacuum and stable conditions, could be more conducive to preserving working tech.

FarmMarsNow
2014-Aug-12, 02:23 PM
AI is the piece that is most likely to become available next. It would be hard to replicate an AI capability, so let us presume that the AI part is not self-replicating but can manage multiple self-replicating units. If you get some proper AI going (and I think the AI we have may be good enough) then the technology exists to set up a self replicating manufacturing system for a particular kind of environment with a known set of resources such as a particular desert with a given kind of sand and other minerals and a predictable amount of light. Then you can use one AI unit to monitor many self replicating manufacturing systems. Each replicating unit doeesn't have to be 'Intelligent'. It just has to be able to work with AI to recognize and process resources in a specific environment with a given set of raw materials, and it has to have a limited set of procedures to purify and use those raw materials. Without requiring the replication of AI and with solid AI in partnership managing multiple units, you get a much easier to design self replicating system.

Noclevername
2014-Aug-12, 04:41 PM
There is some research actively being done to produce self replicating technology, such as RepRap (http://en.wikipedia.org/wiki/RepRap_Project), an evolving 3D printer design that can reproduce most of its own parts.

Jens
2014-Aug-13, 01:06 AM
There is some research actively being done to produce self replicating technology, such as RepRap (http://en.wikipedia.org/wiki/RepRap_Project), an evolving 3D printer design that can reproduce most of its own parts.

It's interesting, but still it is a long, long way from actual self-replication. The wondrous thing about life is that it can take things from an environment and self-replicate using just the natural environment around itself. The machine that's being developed requires certain plastics to be fed into it. Now having a machine that could take minerals from the ground and process them into self-replication would be really cool.

Noclevername
2014-Aug-13, 01:32 AM
It's interesting, but still it is a long, long way from actual self-replication. The wondrous thing about life is that it can take things from an environment and self-replicate using just the natural environment around itself. The machine that's being developed requires certain plastics to be fed into it. Now having a machine that could take minerals from the ground and process them into self-replication would be really cool.

Very true. I can see some kind of chemical processing robot sniffing out carbon compounds to turn into plastics or binding agents for soil particles. Recycled junk can be used for metal, or organic conductors might be applied to a substrate. Maybe it could harvest energy from local ambient sources; I can see a swarm of bots sitting on a hillside waiting for the sun to come up, or fighting for the best spots on a windy rise...

ravens_cry
2014-Aug-13, 04:01 AM
Baby steps. Technology builds on technology. We know self replication is possible, we live it every day. From sunlight, air,and water with traces of the earth, comes every living thing. I sometimes get this weird feeling when looking around and even at myself just how weird that is.

Noclevername
2014-Aug-13, 09:37 AM
Baby steps. Technology builds on technology. We know self replication is possible, we live it every day. From sunlight, air,and water with traces of the earth, comes every living thing. I sometimes get this weird feeling when looking around and even at myself just how weird that is.

A timely Dinosaur Comics strip on the topic: http://www.qwantz.com/index.php?comic=2678

Noclevername
2014-Aug-18, 09:59 AM
There's a possibility of DNA based machines being studied now; I wonder where the lines and overlaps will develop between DNA based technology, genetic engineering, and nanotechnology?

ravens_cry
2014-Aug-18, 05:33 PM
I think we are going to see a lot of the same solutions for quite some time.
Life has had a couple billion years to puzzle at much of the same questions out, so it doesn't surprise me that there should be a lot of wisdom to be gleaned from the nanomachinery of life.