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

GitHub Copilot

copilot.github.com

Post reply on HN
Print view View original

101–110 of 1001 posts

Re: GitHub Copilot

#101

Post by sanderjd »

pjfin123 wrote: post #56
iwintermute wrote: post #21

So if it was trained using "source code from publicly available sources, including code in public repositories on GitHub." was it also GPLv2? So everything generated also GPLv2?

I think this would fall under any reasonable definition of fair use. If I read GPL (or proprietary) code as a human I still own code that I later write. If copyright was enforced on the outputs of machine learning models based on all content they were trained on it would be incredibly stifling to innovation. Requiring obtaining legal access to data for training but full ownership of output seems like a sensible middl…

Certainly not. If I memorize a line of copyrighted code and then write it down in a different project, I have copied it. If an ML model does the same thing as my brain - memorizing a line of code and writing it down elsewhere - it has also copied it. In neither case is that "fair use".

Re: GitHub Copilot

#102

Post by shmageggy »

smasher164 wrote: post #79

I wonder how much of this is OpenAI-based vs program synthesis techniques.

In the FAQ they state that it sometimes outputs non-running or even malformed code, so it looks like fairly pure language modeling with little to no program synthesis.

Re: GitHub Copilot

#104

Post by bravogamma »

Generating new code is not very hard for humans. But maintaining, extending, debugging code - often that’s where the real challenges are. Will the AI copilot be able to fix bugs in its code based on bug reports?

Re: GitHub Copilot

#106

Post by f38zf5vdt »

This could be the kind of thing easy to train and run locally on a GPU.

- You only need code relevant to your language.

- The amount of unique code in your language is going to be relatively small in compared to say, the entire history of internet comment.

- Training a model also shouldn't take long.

I personally would never use an online version feeding all my code back to a home server somewhere, and leaving breadcrumbs that might suggest I'm violating the GPL.

Re: GitHub Copilot

#107

Post by natfriedman »

Asmod4n wrote: post #89
natfriedman wrote: post #60

Hi HN, we've been building GitHub Copilot together with the incredibly talented team at OpenAI for the last year, and we're so excited to be able to show it off today. Hundreds of developers are using it every day internally, and the most common reaction has been the head exploding emoji. If the technical preview goes well, we'll plan to scale this up as a paid product at some point in the future.

Might this end up putting GPL code into projects with an incompatible license?

It shouldn't do that, and we are taking steps to avoid reciting training data in the output: https://copilot.github.com/#faq-does-github-copilot-recite-c... https://docs.github.com/en/early-access/github/copilot/resea...

In terms of the permissibility of training on public code, the jurisprudence here – broadly relied upon by the machine learning community – is that training ML models is fair use. We are certain this will be an area of discussion in the US and around the world and we're eager to participate.

Re: GitHub Copilot

#108

Post by fabiospampinato »

fzaninotto wrote: post #64
fabiospampinato wrote: post #53

Earlier quoted context omitted.

How big/complicated are the functions Copilot is autocompleting for you? I'm thinking perhaps reading 10 potential candidates is actually slower and less instructive than trying to write the thing yourself.

It shows the suggestions line by line, and only shows the best guess. It's not more intrusive than Intellisense. You can actually see all the code blocks Copilot is thinking about if you want to, but that is indeed a distraction.

The problem I see with that is that's not possible for it to understand well which code is the best, GPT-3 is trying to mimic human writing in general, the thing is most human code is garbage, if this system was able to understand how to make code better you could keep training it until you had perfect code, which is not what the current system is giving you (a lot of the times anyway).

Re: GitHub Copilot

#110

Post by qchris »

iwintermute wrote: post #21

So if it was trained using "source code from publicly available sources, including code in public repositories on GitHub." was it also GPLv2? So everything generated also GPLv2?

This is a bit tricky, because at least in the U.S., I don't believe it's settled question in law yet. Some of the other posters on here have said that the resulting model isn't covered by GPL--that's partially true, but provenance of data, and the rights to it, definitely does matter. A good example of this was the Everalbum ruling, where the company was forced to delete both the data and the trained models used they were used to generate due to lack of consent from the users from whom the data was taken[1]. Since open source code is, well, open, it's definitely less a problem for permissively-licensed code.

That said, copyright is typically generally assigned to the closest human to the activation process (it's unlikely that Github is going to try to claim the copyright to code generated by Copilot over the human/company pair-programming with it), but since copyleft in general is a pretty domain-specific to software, afaik the way that courts interpret the legality of using code licensed under those terms in training data for a non-copyleft-producing model is still up in the air.

Obligatory IANAL, and also happy to adjust this info if someone has sources demonstrating updates on the current state.

[1] https://techcrunch.com/2021/01/12/ftc-settlement-with-ever-o...

Post reply on HN