Hacker News Reader

Comment by invig | Hacker News Reader

What's with the default to "if it's not explicitly legal, it must be illegal"?

Imagine if every new piece of software your wrote had to be tested for legality because you don't know that it's explicitly legal. Oh there aren't laws for this new thing, so I guess you should challenge yourself all the way to the supreme court?

I get the author not liking Copilot, but I don't see that GitHub/Microsoft have any kind of obligation to figure this out just because they're GitHub/Microsoft.

If I as an individual had this obligation placed upon me I'd just never write any more code.

Ultimately I think, like open source, Copilot and the tools that will follow advance human progress in novel ways. Software getting easier to make is a good thing. If you don't like this particular implementation of something helpful, feel free to start an open source alternative without challenging yourself in the supreme court.

View on HN

> What's with the default to "if it's not explicitly legal, it must be illegal"?

That's not how I interpret what's happening.

People who produce things have rights over their products. Be it artists, craftsmen, inventors, entrepreneurs or coders. There is a legitimate question here as to whether CoPilot has infringed upon those rights. I don't see it being about "making something illegal." I see it about answering a valid question as to whether CoPilot is liable for measurable damages caused to creators under existing laws.

A few snippets of code is not a product. If there was an open-source money-making product and someone builds a competing product using considerable help from CoPilot then that is a stronger case for damages then if someone just used some snippets of code in their own product.

But at that point, it would be just like someone cloning the Github code without following the license and in that case, it should become obvious that there is a clear violation harming the creators. But in most use cases of CoPilot, where-in people are just using it to build their own product, I doubt there is a cause for damages.

Music samples are a natural parallel.

You cannot sample music without permission no matter how short the sample may be.

Similarly you cannot steal a snippet of someone else's code without permission or the correct licensing.

Eh music sampling is a unique case because of the dual intellectual property concerns (the composition and the recording).

I can’t use a snippet from a recording no matter how short but I can use a tiny snippet of a composition. You can’t copyright a single note.

So code and mechanical music (i.e. a recording) aren't exactly the same, but code and music compositions are more similar. Can you use a tiny snippet from a composition and still infringe copyright? Yes, you can. Maybe not however short, but there will come a point.

> You cannot sample music without permission no matter how short the sample may be.

Which is a blatantly mistaken court ruling and one which I will not enforce if I am on a jury in such a trial.

I'm using the word "product" to mean "something which was produced." I searched for a few definitions because I thought they might actually be different forms of the same word. Turns out I could be wrong about that, but that was my intent. Something that you produce is a "product" of your time, effort, labour etc. Doesn't matter if it's something that you are selling or not. Doesn't matter if it's a relatively small production. The point is you produced it. It is yours.

The question is whether the courts will find damages. Everything else has nothing to do with my comment. You might have your own ideas and opinions, which is fine. So do I. Both are irrelevant. The point is that there is a legal question here that the courts alone are equipped to answer.

a few snippets of a book also isn't a product, and yet it can absolutely be infringing.

If someone bought a copy of an educational book like "Learning Go" and used some code snippets from it in their own product, that's fair use. But if someone released the book as their own product titled "Learning Go Better" then that is a clear violation.

For open-source projects, CoPilot is in the realm of fair-use for snippets but it can be mis-used just like Github can be mis-used if someone blatantly copies a repository.

Fair use only applies in certain contexts, of which writing commercial software is not one of. 'Snippets' are not fair use and it is shocking how many people here think they are.

People get snippets from Stack Overflow all the time and usually there is no concern whether it is properly attributed or not.

I'd argue that people who open-source code expect other people to learn from it in a small way of snippets and that constitutes fair-use.

Code in Stack Overflow answers is explicitly licensed under a permissive license:

https://stackoverflow.com/help/licensing

Copying code samples from a copyrighted work for use in a commercial product is not fair use.

>CoPilot is in the realm of fair-use for snippets but it can be mis-used just like Github can be mis-used if someone blatantly copies a repository.

Important difference is you don't know where your Copilot snippets come from.

If X and Y, then my point is valid!

But if either of those are not true, then the GENERAL point is true and your point is not.

---

Gais, gais, I downloaded the code using an automation tool called a browser, so it's fair use and not infringing!

yay for technicalities!

As a software developer, we deal with specific points rather than generalities. Am I doing something illegal or is it fair use? That's the important point to keep in mind and I assume most of Hacker News audience are software devs.

You also deal with the real world and real people, neither of which operate in binary.

Plenty of people over the years have attempted to skirt the law by putting a proxy in the middle and as such, plenty of clarifications have happened that it doesn't matter.

MS's stance here is specifically that it's the responsibility of the person/company using copilot to ensure the code isn't infringing, it is NOT their stance that the code itself is not infringing.

Their stance is that using it as TRAINING DATA is fair use, so they themselves hold no liability, only their users.

---

It's similar to chicken factories claiming they hold no liability for employing illegal immigrants because those illegal immigrants are the ones who chose to work there. And that they also hold no liability if they go through a 3rd party that exclusively hires illegal immigrants. The law very clearly refutes both stances.

Its better to know, even if you like Copilot and want it to continue.

> I get the author not liking Copilot, but I don't see that GitHub/Microsoft have any kind of obligation to figure this out just because they're GitHub/Microsoft.

Because its a trillion dollar company with an infinite amount of lawyers and legal resources?

> What's with the default to "if it's not explicitly legal, it must be illegal"?

Authors have explicitly and deliberately made it illegal for a person (or corporation) to do what Copilot is doing. Doing it through the legal non-entity of an AI changes absolutely nothing; it's still illegal. To say otherwise is to say that "AI-washing" can be used to nullify any law, which is of course totally absurd. The assumption you lead with is not what anyone is actually trying to argue.

Code on GitHub is "all rights reserved" by default, which means that it is illegal to copy by default. Only by adding a license does it get opened up and made legal to copy - though typically, only if you also include that license when you do so.

So if you trained yourself to only regurgitate github code with wanton abandon and careless disregard for licensing, then yeah, you're liable to violate that default copyright, and certainly going to be violating license rules if you're regurgitating large blocks from memory but not their accompanying licenses.

This is the system that github and Microsoft participate in and willingly and purposely perpetuate. They benefit immensely from copyright law and protection of their code. You can get that they will damned-well avoid letting copilot anywhere near Windows' source, and they would very much enforce their copyright if copilot was spitting that code out for the masses to use.

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.

> What's with the default to "if it's not explicitly legal, it must be illegal"?

This struck me about the article, it's a really bad argument to point out Microsoft didn't point out the law that explicitly made copilot legal. Something not being explicitly legal doesn't make it illegal, although I'm not sure how the courts will rule on copilot.

Reply on news.ycombinator.com