Justice Department officials are turning to the 3D-printing industry to help stop the proliferation of tiny pieces of plastic transforming semi-automatic weapons into illegal homemade machine guns on streets across America.
Taking this purely as an engineering task, how is this remotely possible? I can barely begin to imagine how restrictions on what can be printed could be set. Am I missing something obvious? Some kind of contextual understanding of the object seems to be necessary… please don’t tell me their proposed solution is AI.
In any case it will never work because 3D printing is so easy for makers to do from scratch, so any solution will fail to prevent printed guns from being made.
Again, this is just the pragmatic engineering angle. Please don’t respond with political arguments.
Just spitballing but you’d have to align the desired shape somehow, perhaps with a singular value decomposition. Once its transform was normalized you could compare its shape, or perhaps its convex hull, with a database of banned shapes.
The problem is this is pretty easy to defeat (by adding extra sprues and spikes to the object, breaking it into two shapes, etc) and the more aggressive you get with the check the more you risk false positives.
An AI training set would involve creating a dataset of all the banned shapes, then generating tens of thousands of permutations of them however you believe people might try to trick it. Ultimately the AI would lock onto some small feature of the shape that scores it as positive, perhaps something trivial. That also leads to weird false positives. This also creates an arms race as people figure out what that feature is subvert it.
This problem is much harder in 3D than in 2D (currency). Since you can also cut, file, and glue shit that comes out of a 3D printer later I don’t think this is a solvable problem. Like most gun control measures in the USA it appears to be aesthetics.
You could also just aggressively go false positive all over the place and say “fuck the users”, with exceptions for cops. This is basically the USA’s approach to drones.
Since the best available firmware is open source I don’t see any way of imposing limits on it.
The printer itself doesn’t even know what it’s making since it’s reading directions one by one, so any limits would need to be implemented at a slicer level, which are also basically all open source (at least any worth using).
The only way I could see it working would be mandating that all printers sold in the US come with software checks against it and be non reflashable, but considering a new driver board that would be able to drive 95% of printers is about $25 it is nothing more than screaming into the void.
Which is a very easily recognized pattern, color, and size. The entire point of a dollar is that every single one looks identical.
Imagine if every single dollar bill was a different color, shape, size, printing pattern, etc… Now imagine trying to block that. Now consider that as soon as you figure out how to block all of the current versions, anyone in the world can just design a new version in 5 minutes.
True, but nothing else looks like money. Lots of things have a similar shape as the barrel of a gun.
Money is also quite detailed, with a known list of configurations. Any counterfeit would need to match the details in those known configurations extremely well. Finding that match with a high degree of accuracy is a fairly well understood and common engineering task. This is not the same task as identifying anything that could possibly be used to represent money with a high degree of accuracy, which is essentially what would be needed in the gun printing problem.
Most currencies have a special pattern that printers are programmed to detect and refuse to print. Since illegal gun part designs can’t be forced to include a marker declaring that they’re gun parts, a 3d printer would have to 1) know what a gun is, 2) know how a gun works, 3) be able to tell whether any particular shape could be used as part of a gun, and 4) be able to tell whether any particular shape could be cut and reassembled into a shape that could be used as part of a gun
that’s different, bank notes follow the same pattern/design, the components that could be printed for firearms vary so much in shape and size, even for the same components across different platforms.
Taking this purely as an engineering task, how is this remotely possible? I can barely begin to imagine how restrictions on what can be printed could be set. Am I missing something obvious? Some kind of contextual understanding of the object seems to be necessary… please don’t tell me their proposed solution is AI.
In any case it will never work because 3D printing is so easy for makers to do from scratch, so any solution will fail to prevent printed guns from being made.
Again, this is just the pragmatic engineering angle. Please don’t respond with political arguments.
Just spitballing but you’d have to align the desired shape somehow, perhaps with a singular value decomposition. Once its transform was normalized you could compare its shape, or perhaps its convex hull, with a database of banned shapes.
The problem is this is pretty easy to defeat (by adding extra sprues and spikes to the object, breaking it into two shapes, etc) and the more aggressive you get with the check the more you risk false positives.
An AI training set would involve creating a dataset of all the banned shapes, then generating tens of thousands of permutations of them however you believe people might try to trick it. Ultimately the AI would lock onto some small feature of the shape that scores it as positive, perhaps something trivial. That also leads to weird false positives. This also creates an arms race as people figure out what that feature is subvert it.
This problem is much harder in 3D than in 2D (currency). Since you can also cut, file, and glue shit that comes out of a 3D printer later I don’t think this is a solvable problem. Like most gun control measures in the USA it appears to be aesthetics.
You could also just aggressively go false positive all over the place and say “fuck the users”, with exceptions for cops. This is basically the USA’s approach to drones.
That would be an even more interesting solution for finding new gun-designs for mass-manufacture. Kalashnikov, Winchester, Glock, and Colt watch out!
Since the best available firmware is open source I don’t see any way of imposing limits on it.
The printer itself doesn’t even know what it’s making since it’s reading directions one by one, so any limits would need to be implemented at a slicer level, which are also basically all open source (at least any worth using).
The only way I could see it working would be mandating that all printers sold in the US come with software checks against it and be non reflashable, but considering a new driver board that would be able to drive 95% of printers is about $25 it is nothing more than screaming into the void.
You can also build a 3d printer from scratch pretty easily. Would need to regulate random electronics and robotics components
Open source firmware doesn’t mean anything as long as tivoization is happening.
Which I don’t know whether it’s the case, but legislature might make this a requirement.
I don’t know the answer to the question, but paper printers cannot print bank notes apparently
Which is a very easily recognized pattern, color, and size. The entire point of a dollar is that every single one looks identical.
Imagine if every single dollar bill was a different color, shape, size, printing pattern, etc… Now imagine trying to block that. Now consider that as soon as you figure out how to block all of the current versions, anyone in the world can just design a new version in 5 minutes.
Somewhat related, the US Gov provides play money that you can print for your kids, which I found helpful to teach my kids about how money works. https://www.uscurrency.gov/sites/default/files/download-materials/en/Printable-Play-Money.pdf
That’s honestly kinda funny but also very useful!
True, but nothing else looks like money. Lots of things have a similar shape as the barrel of a gun.
Money is also quite detailed, with a known list of configurations. Any counterfeit would need to match the details in those known configurations extremely well. Finding that match with a high degree of accuracy is a fairly well understood and common engineering task. This is not the same task as identifying anything that could possibly be used to represent money with a high degree of accuracy, which is essentially what would be needed in the gun printing problem.
Most currencies have a special pattern that printers are programmed to detect and refuse to print. Since illegal gun part designs can’t be forced to include a marker declaring that they’re gun parts, a 3d printer would have to 1) know what a gun is, 2) know how a gun works, 3) be able to tell whether any particular shape could be used as part of a gun, and 4) be able to tell whether any particular shape could be cut and reassembled into a shape that could be used as part of a gun
that’s different, bank notes follow the same pattern/design, the components that could be printed for firearms vary so much in shape and size, even for the same components across different platforms.