MIT's magazine Tech Review
generally does better than this, I promise. But in their story on Nathan Myhrvold's recent patent on DRM for 3D printing, they really got it wrong
Let's start with the facts: the patent (#8,286,236 in case you want to look it up) is for a copy-blocking system (DRM) for (some) files used by (some) 3D printers. It will no more "prevent 3-D printer piracy" (by which they mean the use of a 3D printer to create an object that might have IP protections) than any other DRM system has prevented copying. Which is to say, not at all.
The article is full of misstatements such as "you can’t generally copyright objects (exceptions include sculptures and architecture). That’s because copyright applies to creative works but not to 'useful articles.'" Like, say, books? Useful objects, those books. Also, copyrighted. So the article misunderstands what DRM does, misunderstands basic intellectual property concepts, and generally is sub-par.
That aside, what of the patent? The patent works like every other DRM system - you load a file into a DRM-encumbered machine, which checks that "an authorization code" (in the words of the patent) is available allowing you to load this file on this machine to produce some object. That's not a horribly stupid way of stopping people stealing your shape files and making unauthorized use of them - again, just like every other DRM system in existence. But it has jack-all nothing to do with the reproduction of IP-protected objects.
I could just run that file on a machine that doesn't bother to check for the code, or that has a simple resistor soldered in place to ensure that it thinks it has the code for every file. Or I could just take a copy of the file with the DRM stripped out and use that. Or I could make my own file that produces the object in question. Or, or or. I'm sure anyone with more than five minutes to think about it could vastly expand that list of ways this patent is irrelevant to any real problem.
The patent itself has significant problems. Greg Aharonian (of PATNEWS) sent out a mailing earlier this week in which he cited a major government publication that anticipates this patent by five years, as well as several easily-found patents that involve uncited prior art. Just search the US Patent database for patents containing the words "print" (or "printing") and "rights management." I'm sure you'll find the same patents Greg found. Technically, those patents talk about 2D printers, but the problems and solutions they discuss are nearly concept-for-concept identical with this one.
There remain significant problems with how to handle intellectual property protections in a world where anyone can duplicate objects more or less the way we can photocopy book pages. But this patent and article aren't contributing to the discussion.