The thing that caught my eye wasn't so much the specifics as that figuring out how to replicate a thing is much harder than it sounds.
3d printing is a matter of exploring a tiny fraction of a huge realm. I think 3d printing is massively cool, and I'm delighted when I find out about another material it can handle. It looks like a combined additive and subtractive printer exists but costs two million dollars.
A very casual search suggests that there are no 3d printers which can handle more than one material for the same project.
Imagine a device which could examine something and figure out how to make it using conventional manufacturing methods if necessary. I'm not sure how close to a general artificial intelligence it would need to be, but it's hard in a different direction than the research I've been hearing about. I'm sure that partial solutions would be developed long before there was anything approaching a general ability to figure out replication.
In the interests of avoiding the easy science fiction plots, it's just able to replicate things people can make, not living things.
It might be able to do research as well as figuring out how to replicate things by studying them and using knowledge of science and engineering. It wouldn't surprise me if some of that machinery for making vinyl records is abandoned in warehouses or factories.
First link thanks to andrewducker.
This entry was posted at http://nancylebov.dreamwidth.org/1081287.html. Comments are welcome here or there. comments so far on that entry.