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

501–510 of 1001 posts

Re: GitHub Copi­lot inves­ti­ga­tion

#501

Post by mgraczyk »

culi wrote: post #494
mgraczyk wrote: post #477

All this discussion of legality is interesting to me, because I'm pretty sure that if Github ran a search in the background, found the corresponding license for the code snippet, then showed it to the user in some cookie-banner like annoyance, it would be completely legal. This is what Github already does on their website with a search bar. Yet somehow I think most people upset about Copilot would not like that outco…

How do you run a search on the product of a bunch of neural network weights? Do you just mean like double-checking that the code it produced isn't a direct copy of something copyrighted? If so I think that introduces a lot of other issues. There's an interesting phenomenon of different independent comedians suing talk shows for stealing their jokes. It almost always turned out that those jokes weren't actually stolen…

You run search on the text output. Yes I mean that they would search the code. It's certainly legal to show LGPL code snippets along with the license (It may also be legal without this, IANAL).

Re: GitHub Copi­lot inves­ti­ga­tion

#502

Post by AmericanChopper »

ptmcc wrote: post #500
AmericanChopper wrote: post #137

Earlier quoted context omitted.

The Luddite reaction to copilot is very hilarious to me. It seem to be a great way to identify low-talent coders, because who else would possibly feel so threatened by an AI?… Watching HN commenters suddenly become ardent defenders of copyright is quite the sight.

Caring about licenses, fair use, and copyright has been deeply ingrained in the open source and hacker community for literally decades.

Caring about defending peoples right to fair use has been. Which is the exact opposite of the Luddite reaction to copilot.

Re: GitHub Copi­lot inves­ti­ga­tion

#503

Post by mgraczyk »

stingraycharles wrote: post #495
mgraczyk wrote: post #477

All this discussion of legality is interesting to me, because I'm pretty sure that if Github ran a search in the background, found the corresponding license for the code snippet, then showed it to the user in some cookie-banner like annoyance, it would be completely legal. This is what Github already does on their website with a search bar. Yet somehow I think most people upset about Copilot would not like that outco…

It’s interesting and pretty much uncharted territory from a legal perspective, that’s for sure. It very much relates back to the discussion about accountability of machine learning models, in that it’s desirable to be able to explain how/why some output was generated. I don’t think a banner would be sufficient here, though; perhaps some references to the inputs that were used to generate the output, but that’s often…

How could a banner not be sufficient? Under what legal theory is Copilot + banner illegal, but code search is legal?

Re: GitHub Copi­lot inves­ti­ga­tion

#504

Post by andyp-kw »

nightski wrote: post #498
woah wrote: post #491

It would be sad if someone succeeded in shutting down CoPilot for this kind of copyright stuff. It is genuinely useful. I don't care that it reproduces copyrighted content. The only way you can get it to do that is to bait it with the function names of functions that have already been copy and pasted thousands of times onto GitHub without proper licenses. Luckily, someone will probably come out with a "renegade" vers…

Personally I'm not worried about the end user using copyrighted code. That is their responsibility. If you have verbatim GPL code in your commercial closed source code base that is a liability and it might be dangerous to use copilot. What I have more of a problem with is Microsoft charging for copilot which was trained on copyrighted code without any permission whatsoever which they really have no right to utilize/c…

As a human, if I learn how to program by studying copyrighted code, is it unethical for me to use that knowledge to make a living ?

Re: GitHub Copi­lot inves­ti­ga­tion

#505

Post by skydhash »

andyp-kw wrote: post #504
nightski wrote: post #498

Earlier quoted context omitted.

Personally I'm not worried about the end user using copyrighted code. That is their responsibility. If you have verbatim GPL code in your commercial closed source code base that is a liability and it might be dangerous to use copilot. What I have more of a problem with is Microsoft charging for copilot which was trained on copyrighted code without any permission whatsoever which they really have no right to utilize/c…

As a human, if I learn how to program by studying copyrighted code, is it unethical for me to use that knowledge to make a living ?

If you go around pasting code ad verbatim, a lot of people would be upset too.

Re: GitHub Copi­lot inves­ti­ga­tion

#506

Post by jakelazaroff »

woah wrote: post #491

It would be sad if someone succeeded in shutting down CoPilot for this kind of copyright stuff. It is genuinely useful. I don't care that it reproduces copyrighted content. The only way you can get it to do that is to bait it with the function names of functions that have already been copy and pasted thousands of times onto GitHub without proper licenses. Luckily, someone will probably come out with a "renegade" vers…

> It would be sad if someone succeeded in shutting down CoPilot for this kind of copyright stuff. It is genuinely useful. I don't care that it reproduces copyrighted content.

Sure — in the same way that hacking into a competitor's GitHub account and copying their private source code is "genuinely useful" to you. As the person benefitting from unlawfully using their source code, of course you wouldn't care that it reproduces it. But you're not really the person we're trying to help here.

Re: GitHub Copi­lot inves­ti­ga­tion

#507

Post by angusturner »

I get the impression that many peoples' grievance with generative AI (text, code, images etc.) isn't _really_ about the data provenance. Or at least, it feels secondary, compared to the general disruptive nature of the tech.

If tomorrow someone released a StableDiffusion, CoPilot etc with the same functionality, but respecting the provenance of the data (i.e. licensing etc), what concrete difference would this make? Programmers and other creative professionals would still (reasonably) be nervous about the implications for their livelihoods and communities.

At some point it will be possible to prompt a model for music in the style of , and having never heard , the model will generate a convincing emulation, based purely on statistical knowledge gleaned from millions of unrelated songs and text pairs. (I give it 5 years).

Now what? should still be concerned (or not), but at least we're talking about the correct issue: How do we co-exist with generative models that massively disrupt/alter the process of doing creative or intellectual work?

Re: GitHub Copi­lot inves­ti­ga­tion

#508

Post by jakelazaroff »

andyp-kw wrote: post #504
nightski wrote: post #498

Earlier quoted context omitted.

Personally I'm not worried about the end user using copyrighted code. That is their responsibility. If you have verbatim GPL code in your commercial closed source code base that is a liability and it might be dangerous to use copilot. What I have more of a problem with is Microsoft charging for copilot which was trained on copyrighted code without any permission whatsoever which they really have no right to utilize/c…

As a human, if I learn how to program by studying copyrighted code, is it unethical for me to use that knowledge to make a living ?

Unless Microsoft is secretly powering Copilot with Mechanical Turk, that isn't what's happening here.

Re: GitHub Copi­lot inves­ti­ga­tion

#510

Post by stingraycharles »

mgraczyk wrote: post #503
stingraycharles wrote: post #495

Earlier quoted context omitted.

It’s interesting and pretty much uncharted territory from a legal perspective, that’s for sure. It very much relates back to the discussion about accountability of machine learning models, in that it’s desirable to be able to explain how/why some output was generated. I don’t think a banner would be sufficient here, though; perhaps some references to the inputs that were used to generate the output, but that’s often…

How could a banner not be sufficient? Under what legal theory is Copilot + banner illegal, but code search is legal?

I’d argue that copilot’s core functionality is distributing code, which is very different from search (pointing to original code). Very different from a copyright / fair use perspective.
Post reply on HN