The Singularity is Near: From Dust to Device

It’s now possible to scan and replicate an object with moving parts in a 3D printer. Check out what happens when the physicist reaches into the dust and pulls out a wrench. Truly, any sufficiently advanced technology is indistinguishable from magic.

(FYI, I think some post-scanning information must have been added to the computer representation to fully describe the moving parts.)

Hat tip: Kottke.


Interesting, but can the machine print a working miniature version of itself that does the same, essentially simulating infinity?

Post-scan is probably not necessary, and in fact quite difficult if it is to mean what you think. The scanner appears to be detecting non-surface features, including material cracks.

The applications for space travel (and submarines) are substantial.

Yeah, it sounds awesome for space travel. How much energy does it take? Can the raw material be manufactured out of the Martian or lunar regolith?

Space travel would use EBF3
If this takes off, manufacturing would undergo a whole another revolution.

Without post processing how did it know the screw was one whole part when the guy clicks to select red? How did it figure out extents.

X-ray, perhaps, along with some assumption algorithms. Maybe.

So, one more thing to add to your "hand-wave away list" when defending the TGS thesis, right?

Hey -- if people can fabricate most of what they want at home, that would make GDP go *down*! Which must mean our lives are getting worse, because you're incapable of seeing where a measure fails for its intended purpose!

This reminds me of the following comment I saw last year on CalculatedRisk:

This is one of my favorite economy examples. You have an economy of two guys on an island. Tom spends eighteen years and all the island's resources building a huge complicated machine that will manufacture everything in the world, The Transmogrifier. Bob spends eighteen years getting a tan, eating pineapples and sleeping a lot.

How much is Tom's transmogrifier worth?


It's has NO value because.... Bob has nothing to trade for it.

I asked the commenter why Tom would want to sell it, but never got a reply.

In theory the things it makes would also have no value because Bob has nothing to trade for them.

But in practice, Bob has a lot of experience preparing pineapples and that might give him something to trade, plus he is the island's only tanning expert and might sell expert advice about that, etc.

I have it on good authority that capital goods (such as, say, a transmogrifier) can have economic value to their makers, even if they don't sell it, or sell its output, to other people.

That good authority is also know as "common f***ing sense", which seems to be sadly lacking among economists these days...

You only have 30-50 years of life after you're stranded on an island, why waste 18 of them doing anything except getting a tan and preparing pineapples?

I believe that the original quotation erred in that the actual value would be "undefined" since there is no functioning market for it to trade in.

@Colin: People can value a transmogrifier (or its output) over other stuff (where that other stuff requires one to forgo the transmogrifier). Since the transmogrifier is*preferred* to other things, it's clear it holds value, even without a market. And this value will necessarily be well-defined, because there are tradeoffs the maker will and will not make, implying a specific valuation, even if only ordinal.

I agree with your point Silas, but you're missing the point. GDP is a proxy - a rough estimate of economic transactions and well-being. It's a flawed measure, and in the example given a poor measure. But estimators are always poor with few (or no) transactions.

GDP is the best measure we've got, but clearly it misses barter, black and gray markets, and welfare enhancing choices which reduce traded value (such as a working woman choosing to fire her nanny and become a stay at home mom).

But it's not even clear that Mr. Transmogrifier is relatively more happy than Mr. Suntan. That involves interpersonal comparisons of utility. Ex ante they both make utility maximizing decisions, but exogenous shocks (typhoon, cannibals, changing preferences) is more likely to make Mr. Suntan more unhappy.

I think you just figured out Mark-to-Market

This a "dorm room debate" that it at least as old as me - the dorm room was a cave. Our solution was that demand is infinite. Even though Bob is satisfied with his current setup, if presented with enough alternatives he will find one he really wants and will have to come up with a trade to get it - thus an economy. And therefore overall demand can never reach equilibrium. There can be equilibrium for a single product but an economy keeps producing choices, alternatives, price changes so that it is always percolating. And at this point in the debate you're reminded to go out for coffee.

The transmogrifier guy creates a fishing pole and suntan guy trades with fish he catches.

Unfortunately, there is no central bank, so without sufficient money supply they remain in depression.

"I asked the commenter why Tom would want to sell it, but never got a reply."

Because that's how Bob and Tom decided how to calculate eachothers reserve requirements.

Keep trying, Andrew. One day you'll get a laugh.

I remember reading about this technology over 20 years ago, but this is the first instance of application I have seen. Pretty awesome!

Alex posted this, not Tyler.

Oops. Good catch.

Absolutely brilliant!

For all the Alex haters out there, bookmark this and quit bitching:

That is all.

Wow. A machine for making knock-off dildos.

Better than knock-up dildos, I guess...

good one!

Maybe even celebrity dildos if you got one to stand under the scanner.

apropos to that:

It's interesting how this private effort now receives so much attention when a more impressive project at MIT, using 3-d printing to replicate a musical instrument ( ) went by without much notice.

It's interesting here because the input was scanned (even if there was post-processing intervention), rather than meticulously programmed by hand one part at a time.

The latter, which is what the MIT thing is a high-end evolution of, is old boring news to people who've paid any attention to the phenomenon. (Indeed, I know a person who build his own 3d printer from a kit, and produces objects with it the old fashioned programmed-plan way. This is now "ho-hum" territory.)

I've the impression is that 3D-printing makes reasonable copies , but this is not the same as a true duplicate. A printed circuit board reproduced in plastic may look convincing, but without metallic circuits and real resistors, it won't function in your PC. Also, I gather you can pull a 3D-printed copy apart with your bare hands. This isn't Star Trek-style replication yet, in other words. That might come eventually, but we're just getting started.

It depends on the nature of the 3D printer. There are devices that can print arbitrary geometries from sintered steel that are of high quality. Maybe it's not quite the same as a forged aerospace part, but it's pretty good.

Besides, does anyone doubt that mixed-materials printers are just a matter of time? Semiconductor fab already does this on a small scale. It won't be fifty years before these things are churning out working film cameras (for the museum?) and a hundred years before they churn out live frogs.

It won’t be fifty years before these things are churning out working film cameras (for the museum?) and a hundred years before they churn out live frogs.

Actually, apart from the lenses, I'm pretty sure they could do it NOW, with human assembly of the parts. A very basic film camera is easy - people have made pinhole cameras from Legos, and the only reason they don't do lensed cameras is that custom lenses are a royal pain.

(There are custom lens fabrication machines - the hour eyeglass places use them - but I don't know how adaptable they are, and of course they're using VERY specialized optical plastic blanks.

Still, the interesting part is that we can make a film camera from 3d printers right now, if we're either not concerned with lenses or provide them separately.

That said, I bet in 50 years we might be able to make really nice and really complicated ones via a printer. Sure would be nice to have a replica Leica IIIg that didn't cost an arm and a leg.)

The video gives the (false) impression that the printed wrench is the result of the data from the shown scanner demonstration. Alas, the wrench that they print in the video was drawn in CAD; The scanning tool shown is quite cool, but unfortunately cannot produce a working wrench (yet).

I feel cheated. I thought he printed from the scanner coordinates directly.

Also, the red moving screw has to be lined with a support material which is later dissolved in a chemical bath in order to free the joint.

That sort of trick is common in semiconductor manufacturing.

while the video blew MY mind, this is the response from my engineer friend, fwiw:

"I've literally got about 20 3D printed parts on my desk right now. Pretty old tech. The company's printers shown in that video are crap. Good for communication, but nothing functional. If you're researching the space and want to know more we can grab a brew and I'll tell you about some of the better companies."

well, your engineer friend is simply awesome

With all due respect to Arthur C. Clarke, sometimes news is indistinguishable from advertising.

Printer! Tea, Earl Grey. Hot!

It's reasonable to presume that several aspects of this video were very misleading, as follows:

(0) Extremely few items are made from only a few materials, much less one, and most machines have far more than two moving parts.

(1) The printing process took far, far, longer than watching the video.

(2) The moving part, whose boundaries were hidden from the scanner, was added by destructively disassembling a second wrench and scanning the part by hand.

(3) The soft plastic wrench couldn't actually unscrew the tight nut, but instead merely shaved off plastic bits from the wrench.

(4) The oh-so-top-secret plastic composite material is the output of billions of tonnes of chemical plant, employing at least tens of thousands of workers, and would hardly be available in space. By contrast what one can make from sand or other materials actually available on the moon or an asteroid is extremely crude and brittle (see Markus Kayser's "Solar Sinter" project making crude glass bowls out of desert sand using solar power).

Division of labor considerations strongly argue that except for prototyping, and (assuming they can solve the raw materials problem) remote environments like space where transportation from earth is extremely expensive, specialized machines will easily out-compete generalized machines like these 3D printers. In the time it took them to make this fake wrench, Chinese workers made thousands of real ones using a very wide variety of specialized tools and machines.

Ergo, no manufacturing revolution is to be forthcoming from such machines, as interesting as they are. A _design_ revolution, in terms of more rapid prototyping, perhaps, but those Chinese workers and their myriad of specialized technologies are in no danger of being put out of business by people printing wrenches in their homes.

More here:

There is no Great Stagnation.

One. This technology is twenty years old. I bought a "stereolithography" machine that long ago. (See Wikipedia.) Why is this video seemingly going viral this week? What's amazing is that the general public, and apparently economists, remain unaware of this sort of thing for so long. You do know we have like computer controlled milling machines and robot welders and such, don't you?
Two. wrt/ "post-scanning". You mean to say you're so cynical you think they might have cheated just a little for this promotional video? What's the world coming to?
Three. Yes, there are huge differences between this model of the wrench and the wrench. It cannot be used as the wrench would be. (How big a hammer do you think you could hit it with?)
Four. For the purists among you, he was using the wrench model backwards.

Could technology possibly be a life form? As humans, we view ourselves as an organic life form but who is to say technology isn't its own life form that lives off us. We need technology as much as it needs us. I am an avid fan of Kurzweil's theory on singularity.

I guess no one has realised that people have printed a mouse heart on an ink jet printer and got it to start beating.

Also, I'm very skeptical about how much post-scanning technical input was required, given that the printed tool has a ring on it and looks to be slightly different (we didn't see them side by side).

I'm curious how governments will crackdown once IP and weapons will start being produced in mass.

Comments for this post are closed