Let's say you've been working in place A, you show your code to an LLM service (like the dozen or so Copilot-like services) and tell them to refactor. And for the sake of argument, let's say the LLM uses your code and questions for its next training dataset.
A few years pass, then you go to work at Place B, and ask a question that happens to be related to the problem that Place A's code solved, and they give you Place A's code as is.
Incidental to that, I feel like these tools expose the reality behind “copyrighting code/math” and how fallacious it is. If the tool can generate the efficient methods of achieving a result, I think it becomes obvious that one shouldn’t be able to protect it via IP law.
Do you have some evidence that github trained copilot on private repositories? They've been pretty clear about claiming they only used public repos.
Also, gitlab is not owned by Google AFAICT but is instead a publicly traded company.
I think there might be some confusion here between private repositories and public repositories with restrictive licenses. There is evidence of the latter but not the former.
https://devclass.com/2022/10/17/github-copilot-under-fire-as...
Also there are multiple lawsuits already as you've probably noticed.
The claim that was made was that GitHub trained using private repositories and I have yet to see any evidence.
But these kinds of tools can only do that because someone else already put in the work to write the solutions that are used to train their models. Isn't this exactly the kind of situation when copyright is supposed to apply?
If you want to make an argument that math or software shouldn't be copyrighted, LLMs actually make the case for stronger copyright protections.
Maybe, but as long as managers and shareholders all over the world are excited about the upside of the new technology, this is very unlikely to happen. ;)
LLMs would be dead in the water legally, if their owners had to account for every bit of IP the LLMs have been trained with.
Why does that only hold when the result in question is in software? Machines are just tools for achieving results.
The old argument that it's discovered rather than invented is bullshit. Multiple people can always have the same idea for an invention because we think alike and live in the same environment.
Or just ban patents altogether. Of course, this may discourage companies from investing in R&D and that's the real problem: how expensive is it to invent something, and does it justify a 20-year monopoly? But there are no good answers here, and trying to draw a line between math and non-math is bollocks.
Patents exist to incentivize invention. As long as mathematicians are content to do mathematics for the love of it, and they certainly are, there’s no need for mathematical patents.
Practically speaking, mathematical ideas are building blocks not products. Patents on mathematical ideas discourage invention rather than encouraging it because they prevent use of that idea in new products - an idea that would have been discovered anyway. For example the parents of elliptic curve cryptography and arithmetic coding were hugely damaging to invention overall. Patenting a new kind of cork screw doesn’t have this problem, it’s a destination, not an intermediate.
Math can be viewed as a product of how our minds work. We use abstractions to understand and predict the universe, but it's always imperfect, and the theories always incomplete.
E g., you'd think 1+1=2 is some universal truth, except integers don't exist in nature, being just another abstraction that we came up with. And of course, people can rediscover integers repeatedly, but that just says more about how our mind works.
And yes, math is a building block, but so is software. If math theories aren't patentable, that should happen based on them being trivial or perhaps being too useful to society, and not due to some romantic notions of discovery and the universe. Software, too.
So are technological ones.
Here you just mention upsides with very small chance of small downside. See the case for Anthony Levandowski which was much more serious crime as he knowingly created and sold a company with only moat being Wyamo docs and everything bad that could happen to him did happen. He spent 6 months in jail and now he rejoined as CEO in Pronto. The much more probable case is he got to enjoy $680M that Uber gave him and not have to worry about money again.
Maybe he needed all that to actually perform in the new job.
TFA says he had documentation printouts, not code listings. That I can relate to; when learning something new or unfamiliar it's nice to be able to flip back and forth through the physical copy, make annotations, etc.
As far as I know, they were never able to find the Waymo files that Anthony Levandowski stole, because he was at least crafty enough to not load everything directly onto Uber hardware.
"but I took the code and shoved it into my back pocket for a rainy day."
Discussion: https://news.ycombinator.com/item?id=37527720
Also if it was your own code, stealing it to "kickstart" your position at your new employer also feels like a rather bad deal: just bringing your knowledge and spend a good part of your time recreating something you already know how to do sounds like low effort for big money, a much better deal for you if your aim was to get better comp and job security.
The one thing that might skew this equation, and can only be theorized about with the knowledge we have here would be if your new employer DID collude with you and proposed tremendous returns for giving them access - eg a kind of off-the-record deal where you'll engage in corporate espionage & theft of ip for big cash following through some hidden compensation construct.
Assuming the latter would border on conspiracy theories and i don't want to suggest this would have been the case here. Just a play of thoughts to add to the reasoning that YOU doing it on your own for ill-guided hypothetical benefits might not be the sole factor leading to such theft at all.
Maybe the employee sees greater potential at another employer (whether for positive impact on the world, or just better execution of their own vision, or whatever) and wants to pick up where they left off, instead of having to restart.
Everyone seems to assume some explanation involving money or laziness, but it could be for a less selfish (though still illegal/immoral) reason.
While I wouldn’t steal IP, I can appreciate a somewhat similar sentiment: looking forward to patents expiring. There are too many scenarios where a company creates and patents some new invention, only to produce some product that kinda sucks… until the patent expires and someone comes along and does it right and at a lower price.
There's little-to-no personal upside, and only horrible downside if you get caught.
I mean, this guy:
> Moniruzzaman allegedly gave his personal email unauthorized access to Valeo's systems to steal "tens of thousands of files" and 6GB of source code shortly after that development... Valeo said its former employee admitted to stealing its software and that German police found its documentation and hardware pinned on Moniruzzaman's walls when his home was raided.
And while Nvidia presumably hired him for his expertise, they certainly didn't expect him to be stealing code, not even wink-wink-nudge-nudge. Corporate lawyers at trillion-dollar-companies take this stuff super seriously.
So this guy puts himself at massive legal risk... for what? So he can slack off for a few months while he pretends to write code that's already been written -- and gets to browse Reddit? Or so he can deliver code extra-fast in hopes of a quicker promotion -- that may or may not come? Is that really worth it?
It's crazy to me. Why would you risk that?