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

GitHub Copi­lot inves­ti­ga­tion

githubcopilotinvestigation.com

Post reply on HN
Print view View original

961–970 of 1001 posts

Re: GitHub Copi­lot inves­ti­ga­tion

#961

Post by skohan »

A4ET8a8uTh0 wrote: post #842
skohan wrote: post #806

Earlier quoted context omitted.

I don't know why we should be concerned with the status quo of copy-write law at all with respect to AI. ML is categorically new in how it applies to these domains, and it's not clear to me at all that rules that apply to humans have much to do at all with rules which should apply to machines. Imo it is very simple: IP law is intended to incentivize creative work, so that it remains possible to profit from one's crea…

It is an interesting take and it reminds of the thoughts that made crypto what it is today, which made something the lines of: "Old systems suck and our new system is great and it is new technology. Therefore old rules do not apply to it." Not surprisingly, the moment crypto started gaining traction, everyone was quickly made to understand that rules do indeed apply even if it is a new a facet of finance regulations…

I'm not sure if I expressed myself well: I'm not making the argument that we should just accept a future where AI has free rein. Quite the contrary: I would argue we should examine how to preserve the spirit of the precedent - which is to protect creators - which may require us to create new laws.

Re: GitHub Copi­lot inves­ti­ga­tion

#962

Post by abigail95 »

f1refly wrote: post #743
abigail95 wrote: post #564

Earlier quoted context omitted.

That's a problem of the licensers, not for Microsoft or the CoPilot users. If you released AGPL code but never intended to ever sue anyone. Why did you release it like that? If you did and if someone is able to use your code without any damage to you, without reputation loss, and via a way they have access to the innocent infringer defense after you overcome fair use, after you sue them. How is that game over?

So suppose you go out and about and a Microsoft representative punches you in the face. Now, the Microsoft representative has a billion dollar corporation backing him, willing to defend him at all cost through every institution, while you're just John Doe who went on a trip. If you ever went on a hike but never intended to sue anyone. Why did you go out in the first place? If you did and someone is able to punch you…

You have gone off the deep end, please return to sanity.

The court system is the method of enforcement for copyright.

If you want the "right" in copyright, you have to sue people.

To sue people, you need to find infringement. That infringement must be above fair use.

However - if the infringement you find is so minor that you have no loss of revenue or reputation, a court will not award you damages, and may even dismiss the case.

Nobody has any copyright without suing people, there is no copyright police in the general case.

Microsoft has no special rights from its size. Its size makes it a target, it's not beneficial. It's why they have so much trouble with internal rules about GPL. If I infringe on your copyright, the damages will be zero or low, if Microsoft infringes your copyright, the damages could be millions - with the same burden of proof.

Re: GitHub Copi­lot inves­ti­ga­tion

#963

Post by notacoward »

abigail95 wrote: post #724
j2kun wrote: post #205

Earlier quoted context omitted.

The licenses in question in this issue make it explicitly illegal for Copilot to reproduce their code.

I don't care what your license is, I'm going to use it and I'm going to claim fair use. What's explicitly illegal about this?

> What's explicitly illegal about this?

The fact that the work is copyrighted, rights withheld in the absence of a license and limited with one. Also, you should know that willful infringement can carry 5x the statutory damages compared to accidental, and spurious claims of fair use would be distinctly unhelpful to your case. Just by posting that comment, you have probably compromised your position in any future copyright case you might be involved in, or you might even have invited one. I really recommend being more careful when anything legal is involved.

Re: GitHub Copi­lot inves­ti­ga­tion

#964

Post by notacoward »

abigail95 wrote: post #518
bo1024 wrote: post #247

There are two issues -- (1) feeding copyrighted material in to an AI model, and (2) getting copyrighted material out . The latter is obviously a violation of copyright, full stop. The former, to me, is obviously not a violation. If it were, that would massively tilt the playing field in favor of large corporations. It would become very hard to independently train your own models. Philosophically, I go by the principl…

Fair use is baked into copyright law, "full stop". The only way to prevent all uses of your code is to keep it secret. If anyone wants to say me using copilot violates their copyright, then sue me. But if you have no loss of reputation or revenue, and I have an innocent infringer defense - noone can stop me.

> if you have no loss of reputation or revenue

Statutory damages still apply.

> I have an innocent infringer defense

There's no such thing. It's a meaningless phrase, in a legal sense. Claiming that infringement was accidental or unintentional is not a defense. It has no effect on a determination of guilt or innocence. All it affects is the penalty.

Fair use is a defense, but a more limited one than you seem to believe. The usual formulation is "criticism, comment, news reporting, teaching, scholarship, or research" but none of those apply to Copilot. Fair-use claims are also not accepted by default, but only by demonstration that the four factors defining it are all applicable.

Also, since you've brought it up recently, you as the defendant in a copyright case don't get to choose jurisdiction. Usually the plaintiff does, either because it's explicitly defined in the same license that grants anyone rights at all or because it's a place where they do business. If you live in a different jurisdiction that might affect whether the plaintiff or court can collect any penalties, but not whether any are assessed. Having yourself declared persona non grata in multiple jurisdictions doesn't seem like a good long-term choice.

https://copyright.columbia.edu/basics/fair-use.html

Instead of "flooding the zone" with dozens of comments offering nothing but the same few (false) claims - strong echoes of a recently banned user BTW - I strongly suggest you actually read up on copyright and fair use. They're not whatever you want them to be. Courts are unimpressed by your towering intellect.

Re: GitHub Copi­lot inves­ti­ga­tion

#965

Post by asdff »

kamranjon wrote: post #915
tomphoolery wrote: post #905

Does reading software code count as "using software"? I personally don't consider myself subject to a license when I'm reading public code on GitHub. GitHub Copilot and Codex AI seem to be doing nothing more than reading a bunch of source code, not reusing that code to incorporate its functionality into a different product.

Did you read the part of the article where it states that it has been shown that Copilot copies verbatim large sections of copyrighted code, stripped of all attribution, depending on the prompt provided?

I wonder how much code like that floats around stackoverflow

Re: GitHub Copi­lot inves­ti­ga­tion

#966

Post by asdff »

I wish code didn't have any copywrite at all. It should just belong to our species for the benefit of our species. If your entire business model depends on having some private code that you lord over, versus, you know, having some expertise in the field you are in and the ability to generate more code to solve ongoing problems, it seems like you are structured on shaky ground to begin with.

For example, there are plenty of academics these days who are at the tops of their fields and open source all their code. They end up considered as experts not because of a black box code base they implement on problems, but because they can think of potential solutions to the problems at all, and one of the tools used is writing up some code. The code is a shovel or a hammer, its not the one wielding it. They have competitors too of course, just that the secret sauce isn't the code but what goes on in your actual brain.

Its too bad most business leaders fail to understand this, and think its a blackbox code base that makes a decent business. Its the ability to solve problems that matters.

Re: GitHub Copi­lot inves­ti­ga­tion

#967

Post by visarga »

melagonster wrote: post #874
iroh2727 wrote: post #664

Earlier quoted context omitted.

+1. And let's not forget too that "AI", that is, ML models, are not "autonomous" in the way that humans are autonomous. Sure, we use the word "learn" to describe what they do, which is one word that we also use to describe what people do. But ML models are always wielded by people or corporations for particular purposes. If a corporation was to directly publish some copy that appears plagiarized, we'd call that plagi…

There is a common argument that "human just a better neural network", I don't know, did they mean we need to give GPT-3 or Dalle basic human rights?

I am not sure human rights fit the case, but it is one of the most remarkable developments. It is a self replicating distillation of our culture. If our human-based culture grew up and had a baby-culture...

Re: GitHub Copi­lot inves­ti­ga­tion

#968

Post by flkiwi »

Lots and lots and lots and lots of people confusing copyright (an inherent property right granted and protected by the government) and license (a privately granted privilege to use). Butterick—who is no IP fundamentalist, just go look at the license he used for his typefaces—is doing two things: looking at the enforcement of open source licenses so that they are not invalidated by nonenforcement and, related, asking Microsoft to respect the community. I didn’t see him suggest that Copilot is bad or should he shut down, just that they play by the rules. A lot of the reactions here echo a lot of non-developer middle managers who insist that open source code is free and freely usable by anyone for any reason, which simply isn’t the case if FOSS licenses have meaning and value.

Re: GitHub Copi­lot inves­ti­ga­tion

#969

Post by mwigdahl »

elfatizer wrote: post #49

There are lots of comments arguing for or against Copilot on a value judgment, and having an opinion on it being ethical or legal, etc isn't going to be the same for everyone. But I think regardless of where you stand, there should be some sort of legal ruling to clarify the gray areas that Butterick breaks down.

Butterick sneakily asserts over and over that Copilot is simply retrieving code from Github ("Copilot's whizzy code-retrieval methods", "Copilot is merely a convenient alternative interface to a large corpus of open-source code", "our work is stashed in a big code library in the sky called Copilot"). This verbiage seems specifically chosen to present a misleading picture of what Copilot is and does.

Copilot is a set of trained weight values in a matrix. There is no source code stored in that matrix. The fact that someone can prompt Copilot with specifically chosen text to generate a short sequence of code that matches a corresponding segment of code used to train the model does not mean that it is somehow "just retrieving" that snippet. It is _generating_ that code, guided by the weight matrix, via pattern-matching based on the chosen textual prompt and surrounding context.

That distinction is significant because one of the primary defenses against copyright infringement in US law is if the derived work is transformative. Copilot is a work derived in part from Github code, but it has unique capabilities far beyond returning short snippets of input code, and the work itself is clearly an extensive transformation of the input data.

This is without even considering whether concrete _outputs_ of the model that happen to match code in a repository used to train it are themselves protected via copyright or not, which is another issue entirely (and not as cut and dried as many folks on here seem to think).

Post reply on HN