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

281–290 of 1001 posts

Re: GitHub Copi­lot inves­ti­ga­tion

#281

Post by rockemsockem »

lbotos wrote: post #214
rockemsockem wrote: post #194

Earlier quoted context omitted.

I mean, he clearly knew about that code in advance and used his prior knowledge to coax Copilot into spitting it out, yeah? Three characters can get you pretty far, that's 1 combination out of 125,580 (considering all english letters, upper and lower, along with most of the numbers and symbols on my keyboard), plus the description of a fairly complex algorithm. Also, this code is really just executing a mathematical…

> Even if it is a copyright violation, that is one out of, IDK, millions, maybe billions already of Copilot completions? If you, only once, steal lines of code that you don't have license to do so and use them to make money, that's the same exact thing. "Trusting the algo" and saying "whoops I'm sorry" doesn't make a strong legal defense. In a company of 1000 programmers, what are the odds that copilot increases the…

Not all lines of code are made equal under the law. If they were then Oracle would have a copyright on the Java API. Fortunately they do not.

So, no. You can in fact "steal" several lines of code and use them to make money and be legally clean as a whistle. It isn't that clear cut.

Re: GitHub Copi­lot inves­ti­ga­tion

#282

Post by andrewxdiamond »

spywaregorilla wrote: post #268
klyrs wrote: post #163

Earlier quoted context omitted.

> "look I prompted CoPilot for this piece of code that I already knew about and it spit it right out" https://twitter.com/docsparse/status/1581461734665367554 An english description plus three characters of a function name is enough to coax CoPilot into distributing LGPL-licensed code out of context, without a proper license. That's neither "emotional" nor "cherry-picked", it's a clear-cut license violation.

Suppose you wanted to do what some code does, then you see this LGPL code. What can you do? Adjust variable names and play with line spacing and comments until it feels different?

Write it yourself

Re: GitHub Copi­lot inves­ti­ga­tion

#284

Post by notacoward »

invig wrote: post #150

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 ob…

> 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.

Re: GitHub Copi­lot inves­ti­ga­tion

#285

Post by reachableceo »

scarface74 wrote: post #102
smt88 wrote: post #78

Earlier quoted context omitted.

Copilot wouldn't be shut down or neutered because "a few vocal people" protested against it. It would be because it's illegal and violates the licenses, desires, and intentions of the thousands of workers who wrote the code in its corpus. You act like Microsoft is trying to do a public service and people are angry about it. The reality is that they're taking billions of hours of work and using it to build a product t…

> It would be because it's illegal and violates the licenses, desires, and intentions of the thousands of workers who wrote the code in its corpus. I wonder how many people on HN would be on the side of the creators if we were talking about content created by Walt Disney and whether pirating was ethical?

Equal protection under the law.

I am 100% on the side of content creators. Regardless of who they are .

The courts tend to take a dim view of theft. Which is what this is.

The article clearly lays out that multiple requests for sound legal basis have gone unanswered . It simply doesn’t exist and Microsoft is operating on a forgiveness vs permission model.

Licensing is 100% about permissions. Clear and explicit enumeration of the permissions (or lack thereof ) for a work.

This class action lawsuit should surprise nobody. It’s a class that is sick and tired of being exploited.

Do not take my work that I contributed with explicit permissions and use it in a way I didn’t grant permission for. Full stop. It isn’t complicated.

You wouldn’t download a car and all that jazz….

Re: GitHub Copi­lot inves­ti­ga­tion

#286

Post by zo1 »

klyrs wrote: post #163
rockemsockem wrote: post #128

What do people think the future looks like where publicly available resources on the Internet (art, code, etc) aren't fair use for training ML models? Where you have to opt into models or can opt out (and many wind up doing so)? OpenAI, Microsoft, Google, et al will STILL train such models that can do all the same things, but it will be much harder for non-industry-backed individuals to navigate the legal minefield w…

> "look I prompted CoPilot for this piece of code that I already knew about and it spit it right out" https://twitter.com/docsparse/status/1581461734665367554 An english description plus three characters of a function name is enough to coax CoPilot into distributing LGPL-licensed code out of context, without a proper license. That's neither "emotional" nor "cherry-picked", it's a clear-cut license violation.

Here is this guy's function copy-pasted on a SO question:

https://stackoverflow.com/questions/17913191/using-typedef-i...

Found it after 5 mins and a couple tweaks to the search terms.

Another:

https://vdoc.pub/documents/direct-methods-for-sparse-linear-...

Someone copied this guy's book and put it on scribd: https://www.scribd.com/document/514019650/Direct-Methods-for...

Someone put it on a "personal" edu page:

https://people.sc.fsu.edu/~jburkardt/c_src/csparse/csparse.c

A modified version of it here marked as open-source:

https://github.com/rwl/CSparse.py/blob/master/csparse.py

More:

https://tonus.pages.math.unistra.fr/schnaps/schnaps/csparse_...

Google search used to find them:

https://www.google.com/search?q=Sparse+matrix+addition+%22ch...

Could probably find more if I looked harder.

Side note, looks like in a lot of places people do the ""proper""-ish thing and leave this guy's name on the code.

Re: GitHub Copi­lot inves­ti­ga­tion

#287

Post by LASR »

bugfix-66 wrote: post #220
rockemsockem wrote: post #194

Earlier quoted context omitted.

I mean, he clearly knew about that code in advance and used his prior knowledge to coax Copilot into spitting it out, yeah? Three characters can get you pretty far, that's 1 combination out of 125,580 (considering all english letters, upper and lower, along with most of the numbers and symbols on my keyboard), plus the description of a fairly complex algorithm. Also, this code is really just executing a mathematical…

Microsoft Copilot is easily the greatest theft of intellectual property in the history of man. You want to use my code, without ever knowing I wrote it? You want to use my hard work, regurgitated anonymously, stripped of all credit, stripped of all attribution, stripped of all identity and ancestry and citation? FUCK YOU There's no need to defend something so obviously harmful, so why do you do it? The law should be…

Are you saying the act of training the model itself is theft? Or you’re saying that using it is theft?

You can have a totally legitimate business making hacksaws and bolt cutters.

Now if your customers use these tools to break into homes and steal things, then yes, that’s illegal.

But making the hacksaws and bolt cutters is not.

Re: GitHub Copi­lot inves­ti­ga­tion

#288

Post by lamontcg »

cowtools wrote: post #211
rockemsockem wrote: post #194

Earlier quoted context omitted.

I mean, he clearly knew about that code in advance and used his prior knowledge to coax Copilot into spitting it out, yeah? Three characters can get you pretty far, that's 1 combination out of 125,580 (considering all english letters, upper and lower, along with most of the numbers and symbols on my keyboard), plus the description of a fairly complex algorithm. Also, this code is really just executing a mathematical…

Of course it is cherry picked. The idea is that it allows you to INTENTIONALLY void any copyright you want. So let's say I obtain an illegal copy of microsoft windows' source code. Under this precedent, what stops me from just (overfitting) training a neural network to produce the source code verbatim, sans any license notice? But it doesn't end there. What stops me from making a neural network that exactly reproduce…

> The idea is that it allows you to INTENTIONALLY void any copyright you want.

It doesn't void copyright.

Anyone that uses code that Copilot spits out which infringes on someone else's copyright is still liable. There's no requirement for intent. That may be a mitigating factor in terms of remediation, but it cannot void the copyright itself.

It can, however, produce a plague of completely ignorant copyright infringement, and since the user of Copilot has no idea where the code is coming from, there's no way to check if it was trained on infringing code.

If I used Copilot I would be real worried about the legal implications for me and that I could easily be accused of copyright infringement or plagiarism[*]. Of course people seem to not really care these days if they can cheat to get ahead so this is probably a feature, and 99.9% of user won't have their reputations ruined by using it.

[*] Although I'm personally more worried about the fact that most of the code will be wikipedia/blogs-quality and filled with bugs, edge cases and performance issues.

Re: GitHub Copi­lot inves­ti­ga­tion

#289

Post by klyrs »

spywaregorilla wrote: post #268
klyrs wrote: post #163

Earlier quoted context omitted.

> "look I prompted CoPilot for this piece of code that I already knew about and it spit it right out" https://twitter.com/docsparse/status/1581461734665367554 An english description plus three characters of a function name is enough to coax CoPilot into distributing LGPL-licensed code out of context, without a proper license. That's neither "emotional" nor "cherry-picked", it's a clear-cut license violation.

Suppose you wanted to do what some code does, then you see this LGPL code. What can you do? Adjust variable names and play with line spacing and comments until it feels different?

First off, that's a library of pedagogical implementations, so I wouldn't even want to copy it -- I'd prefer a library focused on performance. Second, it's linear algebra, there are alternative implementations and libraries out there. Third, it's covered by the LGPL, so I'd be perfectly happy to link to the library. Fourth, I'd look up a pseudocode description and go from there. In no case would I sit down with another person's implementation and give it the undergrad treatment to pretend that I'm not copying.

Re: GitHub Copi­lot inves­ti­ga­tion

#290

Post by ROTMetro »

pmarreck wrote: post #243
nmilo wrote: post #169

Earlier quoted context omitted.

> What do people think the future looks like where publicly available resources on the Internet (art, code, etc) aren't fair use for training ML models? A better future to me. I don't want pictures of my face training ML models, nor do I want my art, or my code. I don't want my face to be more recognizable to AI, and I don't want my work to contribute to the consolidation of power to a few big firms. And for what, wh…

what if a service could tell you everywhere your photo was on the Internet?

What if a service could scan surveillance videos and show everywhere you have been ever?

Who here is doing a startup to secure licensing rights to every companies surveillance camera videos to make the AI/Surveillance version of Equifax Worknumber? Maybe you offer to give them the surveillance system for free in return for the rights?

Post reply on HN