this post was submitted on 02 Jun 2026
686 points (99.3% liked)

Technology

85168 readers
4794 users here now

This is a most excellent place for technology news and articles.


Our Rules


  1. Follow the lemmy.world rules.
  2. Only tech related news or articles.
  3. Be excellent to each other!
  4. Mod approved content bots can post up to 10 articles per day.
  5. Threads asking for personal tech support may be deleted.
  6. Politics threads may be removed.
  7. No memes allowed as posts, OK to post as comments.
  8. Only approved bots from the list below, this includes using AI responses and summaries. To ask if your bot can be added please contact a mod.
  9. Check for duplicates before posting, duplicates may be removed
  10. Accounts 7 days and younger will have their posts automatically removed.

Approved Bots


founded 3 years ago
MODERATORS
 

This first bill allows the state of California to regulate and oversee all 3D prints in the name of public safety.

you are viewing a single comment's thread
view the rest of the comments
[–] jballs@sh.itjust.works 88 points 3 days ago (8 children)

Under the proposal, printers would have to evaluate STL files, CAD files, or other geometric code using a firearm blueprint detection algorithm and block files flagged as capable of producing a firearm or illegal firearm parts, including conversion devices.

California's Department of Justice, or another relevant state agency, would have until January 1, 2028, to publish performance standards for detection algorithms and software control processes.

This is the problem when lawmakers write technical bills without speaking to technical people. They're going to publish standards for evaluating if your gcode is a firearm or firearm part? THAT'S FUCKING IMPOSSIBLE

[–] SupraMario@lemmy.world 36 points 3 days ago

It's not even that, building a firearm.....is legal...this shit going after printers makes no sense at all, it's fucking legal to print firearm parts.

[–] douglasg14b@lemmy.world 4 points 2 days ago

That's the point. This is just a foot in the door to block your access to print things that might be trademarked copyrighted or affiliated with your corporate overlords.

And a foot in the door to start blocking your right to repair your own things.

Guaranteed.

[–] Folstar@lemmus.org 14 points 3 days ago (1 children)

Fun time to introduce/remind people of the Gell-Mann Amnesia Effect: The phenomenon of a person trusting newspapers for topics which that person is not knowledgeable about, despite recognizing the newspaper as being extremely inaccurate on certain topics which that person is knowledgeable about.

Same thing goes for laws and lawmakers. It's almost as if selecting our "leaders" from a narrow band of society who, for the most part, don't know shit about anything, does not lead to enlightened laws.

[–] jballs@sh.itjust.works 11 points 3 days ago (1 children)

Lol I was just talking to my wife about that yesterday and how it's exactly like AI.

If you read something in the newspaper about your job, you're like "these fucking idiots got this all backwards." If you see AI output of an attempt at your job, it's the same thing.

But if you read an article about someone else's job, you think "that makes sense." Same about seeing AI trying to do someone else's job.

[–] Folstar@lemmus.org 2 points 2 days ago* (last edited 2 days ago)

In most cases, absolutely. There are a few jobs I can see an LLM performing as well or better. Though, then those people would be roaming the streets and do you really want your insurance claim denied while you're out walking your dog or someone yelling about paradigm shifts to align synergies when you're just having a nice day at the park?

[–] prole@lemmy.blahaj.zone 16 points 3 days ago (1 children)

I bet they end up using a fucking llm

[–] theoretiker@discuss.tchncs.de 0 points 2 days ago (1 children)

Yes that's probably how you would do this. Get a bunch of data of gcode of 3d printed gun parts and not-gun parts, for different slicers and printers. Then train some transformer as a classifier. Based on how good object recognition is, i would say its possible that you would get reasonably good accuracy and precision. And because you are scanning for code the architecture will likely be similar to an llm.

[–] isleepinahammock@lemmy.blahaj.zone 7 points 1 day ago (2 children)

Then five minutes later, someone figures out how to make a 3d printable gun that bypasses the gun detector on the 3d printer. It's not like you're printing a whole gun; you're printing parts, most of which look nothing like a gun. How hard would it be to design an algorithm that takes a gun part cad file and then adds a bunch of extraneous pieces to it that can be easily removed? Just keep adding extra crap until the system no longer detects it as a gun part.

[–] theoretiker@discuss.tchncs.de 1 points 16 hours ago (1 children)

Yes that would probably work. There could be some essential features of weapon parts that an algorithm might still be able to learn, and a printer could also keep track of previously printed parts for the classification. I think its unlikely that there are essential features of gun parts that are specific to gun parts so there would probably be a lot of false positives.

Yeah I'm thinking of an automated version of greebling. Except you design the extraneous bits so that they're only attached to the intended print like print supports - something in a non-critical location, easily torn away.

[–] KelvarCherry@piefed.blahaj.zone 1 points 1 day ago (1 children)

just rotate the piece at different angles on the plate, which would change the positions of vertexes and generate an unrecognizable set of printer head instructions. No extra pieces necessary; and even if there were, there's no need for them to be printed attached to the main part.

[–] theoretiker@discuss.tchncs.de 1 points 16 hours ago (1 children)

Isleepinahammocks idea would probably work. But rotation and translation would not. Thats something you can easily take care of in your training data, by reusing the same training data in multiple random positions and random angles.

[–] KelvarCherry@piefed.blahaj.zone 1 points 12 hours ago

are you familiar with 3d printing? The print head's instructions will be completely different if you rotate the model. Unrecognizably different.

[–] SleeplessCityLights@programming.dev 12 points 3 days ago (1 children)

Cosplayers are going to be pissed.

[–] SaveTheTuaHawk@lemmy.ca 2 points 2 days ago

That's a whole section on OnlyFans

[–] Auli@lemmy.ca 3 points 3 days ago

Yes they have no idea what they are asking. Stl is just gcode how do you look for a gun out of coordinates.

[–] KyuubiNoKitsune@lemmy.blahaj.zone 2 points 3 days ago (4 children)

Kinda, render a few images from the gcode, use a CV algorithm to identify the object.

On device it'll be slow or expensive.

[–] M0oP0o@mander.xyz 22 points 3 days ago (2 children)

Define gun is a lot harder then you think. For example

[–] jballs@sh.itjust.works 19 points 3 days ago (2 children)

Your faith in this mystery algorithm is stronger than mine. Here's a diagram of the parts in an AR-15:

So we need an algorithm that renders the gcode I'm printing, then compares it to... something?

[–] ayyy@sh.itjust.works 4 points 2 days ago

Printer: “not a hot dog”

[–] sturmblast@lemmy.world 2 points 3 days ago (1 children)

Theres countless gcode use in the world, much of it is offline

[–] KyuubiNoKitsune@lemmy.blahaj.zone 1 points 3 days ago (1 children)

Doesn't matter. Has nothing to do with online.

You can run OpenCV on an RPi, it's just super slow, and you could probably use a cheap GPU chip to do it faster. You store the pretrained model on the device.

You may even get away with an asic designed for the model, though with that one I'm talking out my ass.

[–] Trilogy3452@lemmy.world 4 points 3 days ago (1 children)

That would makes printers more expensive and my guess is that they'll prefer to force online connectivity