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

621–630 of 1001 posts

Re: GitHub Copi­lot inves­ti­ga­tion

#621

Post by seadan83 »

smt88 wrote: post #390
Aeolun wrote: post #348

Earlier quoted context omitted.

> 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. Yeah, the vocal few. Do you think I give a rats ass that Copilot is duplicating my OS code? I have to imagine most people are completely ambivalent. Of course I have no proof, I just can’t imagine anything else. The lines probably fall somewhere along the MIT vs GPL camps…

> Of course I have no proof, I just can’t imagine anything else. "Ambivalent" means "of two minds," but I'm going to assume you meant that you're indifferent. If people are/were indifferent, their licenses should reflect that. They overwhelmingly don't. Regardless, Microsoft is legally bound to obey the licenses.

I consider GPL as meaning, "don't make the same app as me using this very source code". Cribbing one method out of quarter million line code-base, hardly seems to be redistributing the source code. Literally, it is, but is there no concept scale? Can we go to an extreme and force anyone with a `catch(Exception e)` line in their Java code to go and prove they did not take it from a GPL (or similarly licensed project)? I think this indicates there is a line, at some point it is enough code where you are recreating the functionality of the software - to me that is the thing that matters. I don't give a crap if you use my GPL code to learn from and use any parts of it to build whatever software, but I do care if you recreate the same software using my GPL code.

I would accept a claim of license violation if someone used copilot to autocomplete so many methods from one specific project that you have recreated that original project.

I still think it is a matter of scope. It can still be the case that a relatively small module is not cool to lift, but I think in this case we are still talking about such small subsets of functionality that it is completely divorced from the original software. Like, I could see it if a specific method were really key in some way to a unique application, a very novel solution to a difficult problem - but if that were the case, how can an AI possibly use that for a training model? In other words, the auto-suggestions of an AI are going to be common coding solutions to common coding problems that the AI has seen hundreds of thousands of times. That individual proprietary GPL, unique and novel solution is not really the stuff of an AI suggestion. In other words, the code that co-pilot is going to suggest is going to be non-unique, generic, and not really specific to the overall application at all.

Re: GitHub Copi­lot inves­ti­ga­tion

#622

Post by brigandish »

CoastalCoder wrote: post #165
hn_throwaway_99 wrote: post #106

Earlier quoted context omitted.

Agreed, but I also hate how so much of our substantive law basically has to be created by the courts because (a) many of our legislatures, especially at the federal level, have become more and more non-functional, and (b) IMO legislatures are especially bad at implementing technical legislation. I think there is a good, fundamental legal/societal question of how copyright should apply to AI output. I just don't think…

I agree. Any law that's only clear after a court ruling is, de facto, an ex post facto law. Disgusting.

That's how common law works, it's not disgusting, (unless perhaps you're an overzealous adherent of civil law) nor is it ex post facto. Legislation is produced, (claimed) grey areas are challenged in court, if the outcomes appear unfair then legislators (should) update the law.

Badly written law and poor legislators are a problem in any system.

Re: GitHub Copi­lot inves­ti­ga­tion

#623

Post by truth_seeker »

I am feeling very greedy but ....

With all that intelligence if GitHub Copilot can't produce easy to use and manage full stack framework yet with distributed database inbuilt in either any existing programming language or perhaps a new one created by itself then its not useful for me.

Re: GitHub Copi­lot inves­ti­ga­tion

#624

Post by Siira »

schoen wrote: post #486

Here are a few thoughts I haven't formulated before: It seems clear enough to me that training AIs on copyrighted works is typically or commonly a fair use under existing law, because the AIs can and commonly do learn non-copyrightable elements and aspects of those works. It's very obvious from enormous numbers of examples that current AI systems are capable of learning much more abstract features of human culture (g…

Copyright no longer (and perhaps never) serves as a tool to further the creative/productive output of the society. It should be demolished and rewritten, and when in doubt, it should allow rather than disallow.

Re: GitHub Copi­lot inves­ti­ga­tion

#625

Post by scarface74 »

thrillgore wrote: post #557
scarface74 wrote: post #114

Earlier quoted context omitted.

Yes because it turned out so badly the last time. Microsoft went from being one of the three most valuable companies in the US in 2000 to being one of the three most valuable companies in 2022. Also back then, Microsoft had 90%+ share of the PC operating system market and was bundling IE in its operating system. I’m glad the DOJ forced MS to change its ways.

So they turned around and started buying everyone else. GitHub, Nokia, Activision. They're back to their old shit.

GitHub doesn’t have a monopoly on “hosted git repositories”.

By the time MS bought Nokia, it was already a has been in mobile and the acquisition was a total failure and the game market is competitive.

Re: GitHub Copi­lot inves­ti­ga­tion

#626

Post by neurostimulant »

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…

> I won't be afraid of accidently violating copyright myself, because I won't be trying to bait it into reproducing heavily copy&pasted cherrypicked examples, and I won't use 20 lines of its output with zero modification.

Maybe you can exercise some discipline when using copilot, but what about your coworkers? Many companies might not want their employees to insert copyrighted codes into their projects accidentally.

Re: GitHub Copi­lot inves­ti­ga­tion

#627

Post by a254613e »

>But how will you feel if Copi­lot erases your open-source com­mu­nity?

Jesus Christ, dramatic much? Are people that stumble upon a piece of code while googling how to do something, and end up copying and pasting the code from the repo, really building the open source community? Because that's essentially what it is. Whether I use copilot to generate a tedious function, or I copy it from your open source repo I'm on the same level of being a member of your open source community.

This whole thing feels like artists screaming how AI generated art is horrible, trying to figure out how to sabotage it, or how to start lawsuits - just because their value went down just a bit. Same thing with developers.

Re: GitHub Copi­lot inves­ti­ga­tion

#628

Post by smaudet »

reflexe wrote: post #585
schoen wrote: post #486

Here are a few thoughts I haven't formulated before: It seems clear enough to me that training AIs on copyrighted works is typically or commonly a fair use under existing law, because the AIs can and commonly do learn non-copyrightable elements and aspects of those works. It's very obvious from enormous numbers of examples that current AI systems are capable of learning much more abstract features of human culture (g…

Maybe the solution here is do adopt the approach from humans: if you independently produce someone's copyrighted work and then discover about it - you'll drop it. The same approach could be used here, they can add a check for the similarity between the output and the original training material, if it is above a threshold, they'll drop the suggestion (maybe they are already doing that).

That doesn't work, though - the problem is you have automated crime. When you do this, you can no longer handle this on a case by case basis - you have to resort to automating justice.

And at this point, you are getting both attacked and supported by AI, and not really better off in a meaningful way.

There's no good way of solving this issue, without general intelligence, and the problems that will bring (what reason does a generally intelligent AI have for supporting us or not enslaving us).

This is why all AI research IMO is unethical. Point me to a single AI use that has not already been abused, maybe I will change my mind, as it stands though we should be prosecuting the people misusing this technology, or at least irresponsibly releasing it, as fast and as quickly as possible before we get the point we are no longer fighting bad actors but the machines themselves.

Re: GitHub Copi­lot inves­ti­ga­tion

#629

Post by thomasahle »

trap_goes_hot wrote: post #586
mkr-hn wrote: post #529

A sizable, possibly plurality cohort of fully adult tech people is young enough to not know about United States v. Microsoft Corp. This would explain a lot of comments I see on this topic. If you don't know Microsoft's history, a lot of what more informed people are worried about seems overblown. Copilot was Microsoft's first test of people's trust after the GitHub acquisition. It's going very, very, very poorly. The…

>If you don't know Microsoft's history, a lot of what more informed people are worried about seems overblown. Or maybe they do know about it, and don't agree with you. Do you allow for such an option? https://github.com/features/copilot "What can I do to reduce GitHub Copilot’s suggestion of code that matches public code? We built a filter to help detect and suppress the rare instances where a GitHub Copilot suggesti…

> You have the choice to turn that filter on or off during setup.

Notice that Copilot often gives code that verbatim matches opens source software, even when that filter is on. For example: https://twitter.com/DocSparse/status/1581461734665367554?s=2...

Their approach of "matches or near matches (ignoring whitespace)" is clearly inadequate, and it's honestly insulting that they think this is enough. Even if Copilot just changed the case of a single letter, their filter wouldn't catch it.

Re: GitHub Copi­lot inves­ti­ga­tion

#630

Post by smaudet »

xvector wrote: post #607
chrismcb wrote: post #605

Earlier quoted context omitted.

It seems pretty clear to me, training an AI on copyrighted materials is not fair use. I'm not sure why you seem to think it is fair use

> I'm not sure why you seem to think it is fair use I think OP explains clearly, in many paragraphs, why it's fair use. That's literally what their whole post is about.

The concept of fair use applies to small volumes of work.

Clearly, training on large volumes of data is not small volumes in any sense of the word. The argument that it is fair use is itself flawed.

Post reply on HN