Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

GitHub Copilot and open source laundering

drewdevault.com

Post reply on HN
Print view View original

11–20 of 29 posts

Re: GitHub Copilot and open source laundering

#11

Post by Engineering-MD »

MartijnBraam wrote: post #3
sharemywin wrote: post #2

An interesting question would be over fair use. Generally, "transformative" works are covered under fair use. I can certainly see the argument that DL output is transformative in nature. Not a lawyer take my comments with a grain of salt.

Is plagiarism legal if I change the variable names?

Just for clarity, note that plagiarism and copyright infringement are distinct. If you rewrite someone’s paragraph with synonyms it’s still plagiarism but not copyright infringement especially if you don’t cite the work in question. Conversely, you could quote several paragraphs of someone else’s work in quotations with citation and it may be copyright infringement but not plagiarism.

Or at least that is how I understand it.

Re: GitHub Copilot and open source laundering

#12

Post by thruuavay »

Another approach is to write a program, "bugify", that deliberately inserts bugs into software. Then, create many users and repositories. This would cause the dataset's value drop even lower than it currently is (the value is already low because bugs are everywhere and developers who use Copilot are even more careless than ones who type it all; it's basically a form of regurgitate-and-paste).

Re: GitHub Copilot and open source laundering

#13

Post by jhoelzel »

thruuavay wrote: post #12

Another approach is to write a program, "bugify", that deliberately inserts bugs into software. Then, create many users and repositories. This would cause the dataset's value drop even lower than it currently is (the value is already low because bugs are everywhere and developers who use Copilot are even more careless than ones who type it all; it's basically a form of regurgitate-and-paste).

I wonder how many forks and github users you would habe to emulate to actually impress CoPilot and if they have a scoring system for the commited code.

I imagine that code in the nginx repository is much higher rated then my personal code that I have open sourced.

In psychology reliability and retest quality are important factors so I hope they are for something like this too.

Re: GitHub Copilot and open source laundering

#14

Post by ddevault »

ghoward wrote: post #9

Drew, that is a great post. I have two comments and a question. First, my comment is that GitHub claims the right to use any public code on the Internet, not just GitHub. This means that they claim the right to mine code on Sourcehut, or in my case, on my personal Gitea instance. I'm not sure your suggestion 1 covers that. I think it should. Second, users could fork code from Sourcehut onto GitHub, and then even if G…

I think that your clause is fine and does not cause these licenses to be non-free. But it is worded a little bit weirdly, by stating that the output of Copilot etc is literally "this software" rather than a "derivative work". I might suggest clarification on that part. Cheers!

Re: GitHub Copilot and open source laundering

#15

Post by omegalulw »

sharemywin wrote: post #2

An interesting question would be over fair use. Generally, "transformative" works are covered under fair use. I can certainly see the argument that DL output is transformative in nature. Not a lawyer take my comments with a grain of salt.

> I can certainly see the argument that DL output is transformative in nature.

It's a thin line. Depending on how the model is trained and what size it is, it is not unreasonable to expect that it will have the ability to remember training data.

Re: GitHub Copilot and open source laundering

#16

Post by ghoward »

ddevault wrote: post #14
ghoward wrote: post #9

Drew, that is a great post. I have two comments and a question. First, my comment is that GitHub claims the right to use any public code on the Internet, not just GitHub. This means that they claim the right to mine code on Sourcehut, or in my case, on my personal Gitea instance. I'm not sure your suggestion 1 covers that. I think it should. Second, users could fork code from Sourcehut onto GitHub, and then even if G…

I think that your clause is fine and does not cause these licenses to be non-free. But it is worded a little bit weirdly, by stating that the output of Copilot etc is literally "this software" rather than a "derivative work". I might suggest clarification on that part. Cheers!

That's a good suggestion. Unfortunately, "derivative work" is an established legal term, and was before the GPL, so I don't think I should use it. In fact, it should be part of my definition, now that I think about it.

However, if I can come up with some other term, then I'll definitely take your suggestion. I'll just define it to be "this software," any derivative work, or the application of an algorithm...and then make the license apply to that definition.

I hope that makes sense and is what you meant. If not, please let me know.

Re: GitHub Copilot and open source laundering

#17

Post by PoignardAzur »

> I cannot speak for the rest of the community that have been hurt by this project, but for my part, I would be okay with not pursuing the answers to any of these questions with you in court if you agreed to resolve these problems now.

I mean... has anyone really been "hurt" by Copilot?

Copying from another thread... The concept of intellectual property exists to incentivize creating valuable art/literature/code. In theory at least, we agree to uphold IP laws because we recognize that more value gets created when they're a state enforced monopoly on the person who came up with that piece of art/literature/code.

But we also recognize that sometimes these laws go too far; eg that there are patent trolls and corporations fighting public domain and game publishers going after anyone who makes a let's-play of their video.

In those case, it's reasonable to think the world would be better off if we all shrugged and told the IP holders "too bad, someone else is going to create value off your work and you're not going to get a cent from it, we just think it's not worth building and maintaining a nightmare bureaucracy just so you can tax them".

And from that point of view... Copilot is fine? It's not like the people posting code on Github or StackOverflow were thinking "I'm only doing this because I know a future AI 10 years from now won't scrap the code I wrote to train a neural network to create a code completion engine". Yeah, yeah, this breaks the spirit of the GPL and Stallman's vision, etc, etc.

But... I mean, at some point, you got to stop debating semantics and wonder what we're coding for. What Microsoft has created is a tool that can collectively save developers billions of man-hours. It's a net good for humanity. As far as I'm concerned, the fact that this net good was developed is infinitely more important than the fact that Microsoft didn't pay royalties to a nebulous amount of developers who wouldn't have noticed anything if Microsoft hadn't developed Copilot.

tldr MIT license is great, piracy is great, fanfiction is great, screw the very concept of intellectual property.

Re: GitHub Copilot and open source laundering

#18

Post by ddevault »

PoignardAzur wrote: post #17

> I cannot speak for the rest of the community that have been hurt by this project, but for my part, I would be okay with not pursuing the answers to any of these questions with you in court if you agreed to resolve these problems now. I mean... has anyone really been "hurt" by Copilot? Copying from another thread... The concept of intellectual property exists to incentivize creating valuable art/literature/code. In…

I actually agree, but this is a pretty deep philosophical issue that requires several more blog posts to fully unpack - blog posts I will write.

Essentially, I agree that the very idea of intellectual property is invalid, and my end-game would be to have it overturned entirely, copyright and patents alike banished from this earth. However, a practical view must acknowledge that we live in a world where IP is normalized in the view of the public and in the view of the law.

I would ordinarily "walk the walk" here and act as if the world operated according to my moral lens, i.e. not bugging people about compliance with software licenses. In this case, though, that would be counter-productive. The issue is that there is an asymmetry between the software producer and consumer. I believe that the user has a right to the source code for all software they use (praxis: reverse engineering is good even when it's illegal). However, under the current copyright regime, the producer can withhold their source code. If I publish open source software without concern for copyright (e.g. WTFPL), this allows commercial entities to take my code, but does not allow me to take theirs, so there is a power imbalance that works against my interests.

So, free software licenses are necessary and important. In an ideal world, without copyright, I would be totally fine with GitHub Copilot - because I would also have access to its source code, and the model, and would be able to modify, repurpose, or sell it myself. But we don't live in that world, unfortunately, which allows GitHub to exploit the power imbalance by ingesting free software into a commercial product that they have the sole right to monetize and dictate terms of use for. It's this exact imbalance that the GPL family of licenses were written to address, and it's a spit in their face to pretend that they don't apply, especially from an organization that damn well ought to know better.

Re: GitHub Copilot and open source laundering

#19

Post by lithos »

sharemywin wrote: post #2

An interesting question would be over fair use. Generally, "transformative" works are covered under fair use. I can certainly see the argument that DL output is transformative in nature. Not a lawyer take my comments with a grain of salt.

The US supreme court already believes it's dubious that an AI can create an original work: https://www.google.com/amp/s/www.natlawreview.com/article/pa...

Though has ruled that training AI on copyright materials is fine, as long as the defendants can claim it's a card catalog: https://towardsdatascience.com/the-most-important-supreme-co...

So it's going to be interesting to see how this is viewed by the supreme court when it gets there.

Re: GitHub Copilot and open source laundering

#20

Post by haskellandchill »

torginus wrote: post #8

Just recently I was trying to implement a feature using an extremely complicated and obscure API (some Chrome internal C++ stuff) and the only place I found a working example was inside a GitHub project which was GPL. Obviously if I took the code and modified it to suit my purposes I would've run afoul of the GPL, but if Copilot suggests the same snippet to me it'd be fine. So there's obviously an issue here.

> Obviously if I took the code and modified it to suit my purposes I would've run afoul of the GPL

Do you honestly believe that? I refuse.

Post reply on HN