We do have 3d printers now. There's even the "RepRap" project, aim for the 3D printer to print all its own parts. RepRap - RepRapWiki Main obstacle the computer chip but we have nanoscale 3D printers as well.
Once a 3D printer can make itself from simple stock materials - then you need to find a way to print out all the components for asteroid mining. But some forms of mining are very simple. Particularly, you can extract nickel by surrounding a meteorite in a bag which you don't need to heat very far, by the Mond process. Can heat it by solar power. And end result is nickel vapour which is all ready for use as a 3D deposition printer. So, as long as you can make do with mainly nickel parts, you could get much of the way using just this technology. Of course would need to have other elements as well, it's just a start. But - a bit like a miner making a hammer first - make the easy tools first, using this method - and then - gradually make more complicated machines for harder to extract but important elements.
In principle we could already make a start at a "clanking replicator" using very slow computers printed using existing 3D printers - or even with clockwork "babbage machines"?
But in practice it would be hugely expensive surely, if it did work, a kind of "Babbage machine" type project.
It could be made safe. One way - to just limit the number of generations. If you can only replicate for a dozen generations and stop, say, and you have mechanisms in place that immediately destroy any flawed replicators, they destroy themselves, and also some method that lets you instantly stop the whole thing if anything goes wrong - that's how I'd design it anyway. Have a continuous beacon you have to keep sending, which is encrypted in a non reversible way - so they can't generate their own signal - and design so they all stop working if you stop sending this beacon. For interstellar travel it could be just a beep every hundred years, enough so all the von neumann probes stop working and become inert if the parent civilization forgets about them or becomes extinct.
You'd also need to make sure any such mechanism can't be bypassed by fledgling civilizations that might find the probes and not understand the need for a failsafe like that.
So- yes I think we are not that far away from it, maybe a few decades, but would need to be very careful.
We can of course make self replicators right away if they are living cells, microbes. And so taht's another approach, synthetic life. Which requires a very high level of care if we ever design synthetic life expected to reproduce in the wild. So far any experiments are carefully designed so they can't possibly reproduce outside the laboratory (main one was an experiment that added two extra bases making DNA with six bases - designed so the new bases couldn't be made in the wild, only in a laboratory).
That counts as a von Neumann machine. So long as they replicate, doesn't matter for the definition that they can only replicate in particular environments.
It doesn't have to be an interstellar replicator for instance. If we can make self replicating solar panels that can replicate themselves using materials available in Near Earth Asteroids - that is still a Von Neumann machine.
And any form of synthetic life is a von Neumann style self replicator in its broadest sense, if artificially made.
Here I'm not suggesting that we are at all close to a general purpose self replicator that can survive almost anywhere by prospecting for materials and using creative fluid ways of working with them as in many science fiction stories. Never mind the idea of a self replicator capable of any degree of intelligence.
But we might not be far away, as in, a few decades perhaps, from a self replicator that can make copies of itself in special situations, e.g. to make solar panels from metal rich / silicon rich near earth asteroids using comparatively simple tools designed to work in that situation, and that situation only.
See also Robert Walker's answer to What should be humanity's approach to exploring space and discovering alien races once we're able to travel to other planets?
Once a 3D printer can make itself from simple stock materials - then you need to find a way to print out all the components for asteroid mining. But some forms of mining are very simple. Particularly, you can extract nickel by surrounding a meteorite in a bag which you don't need to heat very far, by the Mond process. Can heat it by solar power. And end result is nickel vapour which is all ready for use as a 3D deposition printer. So, as long as you can make do with mainly nickel parts, you could get much of the way using just this technology. Of course would need to have other elements as well, it's just a start. But - a bit like a miner making a hammer first - make the easy tools first, using this method - and then - gradually make more complicated machines for harder to extract but important elements.
In principle we could already make a start at a "clanking replicator" using very slow computers printed using existing 3D printers - or even with clockwork "babbage machines"?
But in practice it would be hugely expensive surely, if it did work, a kind of "Babbage machine" type project.
It could be made safe. One way - to just limit the number of generations. If you can only replicate for a dozen generations and stop, say, and you have mechanisms in place that immediately destroy any flawed replicators, they destroy themselves, and also some method that lets you instantly stop the whole thing if anything goes wrong - that's how I'd design it anyway. Have a continuous beacon you have to keep sending, which is encrypted in a non reversible way - so they can't generate their own signal - and design so they all stop working if you stop sending this beacon. For interstellar travel it could be just a beep every hundred years, enough so all the von neumann probes stop working and become inert if the parent civilization forgets about them or becomes extinct.
You'd also need to make sure any such mechanism can't be bypassed by fledgling civilizations that might find the probes and not understand the need for a failsafe like that.
So- yes I think we are not that far away from it, maybe a few decades, but would need to be very careful.
We can of course make self replicators right away if they are living cells, microbes. And so taht's another approach, synthetic life. Which requires a very high level of care if we ever design synthetic life expected to reproduce in the wild. So far any experiments are carefully designed so they can't possibly reproduce outside the laboratory (main one was an experiment that added two extra bases making DNA with six bases - designed so the new bases couldn't be made in the wild, only in a laboratory).
That counts as a von Neumann machine. So long as they replicate, doesn't matter for the definition that they can only replicate in particular environments.
It doesn't have to be an interstellar replicator for instance. If we can make self replicating solar panels that can replicate themselves using materials available in Near Earth Asteroids - that is still a Von Neumann machine.
And any form of synthetic life is a von Neumann style self replicator in its broadest sense, if artificially made.
Here I'm not suggesting that we are at all close to a general purpose self replicator that can survive almost anywhere by prospecting for materials and using creative fluid ways of working with them as in many science fiction stories. Never mind the idea of a self replicator capable of any degree of intelligence.
But we might not be far away, as in, a few decades perhaps, from a self replicator that can make copies of itself in special situations, e.g. to make solar panels from metal rich / silicon rich near earth asteroids using comparatively simple tools designed to work in that situation, and that situation only.
See also Robert Walker's answer to What should be humanity's approach to exploring space and discovering alien races once we're able to travel to other planets?
About the Author
Robert Walker
Writer of articles on Mars and Space issues - Software Developer of Tune Smithy, Bounce Metronome etc.Studied at Wolfson College, Oxford
Lives in Isle of Mull
4.8m answer views110.3k this month
Top Writer2017, 2016, and 2015
Published WriterHuffPost, Slate, and 4 moreHuffPost, Slate, Forbes, Newsweek, Business Insider, and Science 2.0