Live data from Hacker News

GitHub Copi­lot inves­ti­ga­tion

githubcopilotinvestigation.com

371–380 of 1001 posts

Re: GitHub Copi­lot inves­ti­ga­tion

#371

It's always interesting to see the buzz that occurs when Copilot is brought up as a topic. This place is called "HackerNews", yet routinely people forget that a "hacker" is somebody using technology to overcome novel problems. Doesn't GitHub Copilot fall into this category? Why is there such an outcry over a technology that has been in the public's hands for less than a year? I'm almost certain that the team responsi…

Nothing is in the publics hands; they have taken the public data and given back nothing but a blackbox that you pay money for.

They don't even trust the thing to train it on their own code, yet their boss is over here telling us they are "learning". It's a damned insult.

Re: GitHub Copi­lot inves­ti­ga­tion

#372
post #310

Earlier quoted context omitted.

No one wants to stop what Copilot is being sold as. They want to stop the company selling it from doing what it is famous for doing.

Copilot is not selling code, they're selling GPU time. If you're ready to buy a hundred GPU/TPUs to train a new Copilot that is just as good, but for free, then go do it please, everyone will thank you

The tagline is "Your AI pair programmer"

That's the pitch. You're renting a pair programmer.

If your pair programmer is stealing code, you're going to have a bad time. This has nothing to do with...whatever you're on about.

Seriously though. Did you click the wrong reply link?

Re: GitHub Copi­lot inves­ti­ga­tion

#373

Does GitHub not have the right to view and train from your content when you agree to their Terms of Service and upload your code? People are conflating their open source license with the one they give GitHub when making a GitHub account, but they are two entirely separate and parallel licenses. The former is for other people to use your code, the latter is for GitHub to host your code. If you don't like it, you are f…

The current github terms of service don't seem to mention this use when they describe the license granted github. https://docs.github.com/en/site-policy/github-terms/github-t... 4. License Grant to Us We need the legal right to do things like host Your Content, publish it, and share it. You grant us and our legal successors the right to store, archive, parse, and display Your Content, and make incidental copies, as n…

If anyone asked me to define Copilot I'd refer back to this:

> parse it into a search index or otherwise analyze it on our servers; share it with other users

That is the most succinct and most accurate definition of Copilot I've ever seen.

Re: GitHub Copi­lot inves­ti­ga­tion

#374

Why would anyone want to stop Copilot is beyond me. Reinventing the wheel, millions of time a day, is an atrocity. Millions of (wo)man hours, wasted, every single day, on writing solutions to problems that have already been solved. There is a partial solution to this, and it's making people angry, it's crazy. If you put your code publicly on the internet, you should expect that people will reuse your code at some poi…

> Reinventing the wheel, millions of time a day, is an atrocity.

> Millions of (wo)man hours, wasted, every single day, on writing solutions to problems that have already been solved. There is a partial solution to this, and it's making people angry, it's crazy.

Following this line of thought, do you think that all code from all software should be open source and publicly available (and free to copy and use), in the interest of saving more person hours from reinventing the wheel?

Re: GitHub Copi­lot inves­ti­ga­tion

#375
post #90

I feel "stealing your community" is lawyer hyperbole, but people also seem ok with what MS is doing with copilot, and I am not. If you think what copilot is doing is ok, and there is nothing wrong with it, I'd love it if you could go through this small thought exercise, and see if it impacts your view at all: Say you write a bunch of code, and release it under GPL. For the sake of argument imagine it is something com…

> Another person is trying to solve this problem in some commercial software. They find your code, copy-paste it into their project, then sell their software and don't release the source, or even acknowledge you. Let's be honest, copy-and-pasting happens all the time. In software, in engineering, in marketing, in everything. Whether people acknowledge it or not. Everyone looks at Stack Overflow all the time. You do i…

> Let's be honest, copy-and-pasting happens all the time. In software, in engineering, in marketing, in everything

If I copied code without the rights to it into code I have written at any company I would absolutely be fired. It would not be up for debate.

> Everyone looks at Stack Overflow all the time. You do it. I do it.

I don't - it is very infrequently that I would look at SO answers

> Nobody reads the licensing terms.

Yes, they absolutely do, because again OSS or the GPL is meaningless if people are ignoring the license. More over I would suggest you talk to your employer's legal and IP departments to let them know you're copying code you don't have rights to into their product.

> We all produce software with reskinned and taped together functions. A collage is still unique, creative, work despite being glued together with other people's art.

Wow.

Absolutely not.

Competent engineers know how to write code themselves, they aren't copy-pasting their way to a solution. That's why you get paid a lot - if I was happy with copy pasta solutions I would hire a bunch of badly performing uni students.

> Most songwriters will write a song with a part like someone else's song.

If two people write similar songs that does not mean one copied the other. If one person copies part of another person's song they will end up in court, and they will lose all the revenue from the entire work.

> The products you buy at a store are rip-offs of someone else's product.

The ripoffs that stay on the market are not made by copying the entire implementation.

> Everybody stands on the shoulders of giants before them. Such is learning, such is life.

I didn't say anything at all to imply that we didn't. What I said was you don't get to just copy other people's work and pass it off as your own.

> Get over it.

Just because you apparently can't actually write new code yourself doesn't mean that that applies to other people.

Also, you should really get your employer to tell you whether your proposal of ignoring copyright is ok.

What you are saying is that if I google some problem, and copy some code from, say, gecko, or linux, or gcc, etc into my proprietary closed source product, that perfectly ok and those silly open source people should keep it to themselves if they don't want me doing so.

But there's also a difference between "I searched for this and copied the code I found" and "I typed some letters

Re: GitHub Copi­lot inves­ti­ga­tion

#376
post #163

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.

It's a violation but this examples doesn't jive with how I use Copilot, for whatever that is worth. Usually Copilot takes into account context and integrates the various bits into a whole that fits in with the surrounding code. The "empty file, type function signature" use case feels somewhat dubious to try to understand actual, not theoretical, harm. Though I respect the right for this person to defend their rights.

Re: GitHub Copi­lot inves­ti­ga­tion

#377
Here's the thing about GitHub that most people do not realize. I find this funny because of all the talk of following license agreements, very few have taken the time to read the terms of service for GitHub.

from their terms of service: "Short version: You own content you create, but you allow us certain rights to it, so that we can display and share the content you post." emphasis mine.

that's what they call the "Short version" of the following paragraphs, which are found here: https://docs.github.com/en/site-policy/github-terms/github-t...

they allow themselves the right to display content you upload to others. GitHub does not seem to really put a cap on that in terms of what intentions it needs to have or for what purposes it needs to share your content.

this seems to me that, by putting your code on github.com, that you are granting GitHub license to show it to others. period. IANAL, but it seems like all code anyone puts on github.com is dual-licensed, at least. GitHub gets their own rights to your code.

I read this before I signed up, and while I can't remember if this exact passage was present at the time, I was ok with everything GitHub wanted at the time, and I continue to be.

githubcopilotinvestigation.com doesn't seem to have much hope of doing anything except getting people mad. but you all were already mad anyway, weren't ya?

Re: GitHub Copi­lot inves­ti­ga­tion

#378

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…

I don't think that having emotional discussions around technology that does no less than provoke homan emotions on command is low quality.

Re: GitHub Copi­lot inves­ti­ga­tion

#380

My view is the copilot is not stealing open source code. It is learning from it just as a human reader would. People's disguste is based on the assimilation of what they thought was a human trait being machine derived from their work. The copilot service backed by an army of actual humans wouldn’t be a story at all. Nor would anyone be angry, if an individual offered coding skills as a service, and had gone through t…

> It is learning from it just as a human reader would I don't see how that invalidates the copyright/license argument. So, instead of just a straight up license violation it's a license violation via plagiarism. That argument wouldn't hold up even if it was a human that caused the violation. You can't just paraphrase someones licensed work and then lie about looking at and pretend you made it yourself, which is basic…

It learns the same way a human does by learning patterns. It is not illegal to comprehend how to accomplish tasks by reading other people's source code.

The arguments against my point always assume perfect memory of everything this model is consumed. This is the plagiarism position. In reality, some patterns are more common than others and generate a code that looks nearly identical. I can’t speak for the reasons for this, as I’m not familiar with all of the methods. However, I don’t assume that is the current working state or intent of Codex.

Post reply on HN