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

901–910 of 1001 posts

Re: GitHub Copi­lot inves­ti­ga­tion

#901

Post by judge2020 »

BrainVirus wrote: post #896
iroh2727 wrote: post #664

Earlier quoted context omitted.

+1. And let's not forget too that "AI", that is, ML models, are not "autonomous" in the way that humans are autonomous. Sure, we use the word "learn" to describe what they do, which is one word that we also use to describe what people do. But ML models are always wielded by people or corporations for particular purposes. If a corporation was to directly publish some copy that appears plagiarized, we'd call that plagi…

A lot of really good points. >Sure, we use the word "learn" to describe what they do, which is one word that we also use to describe what people do. But ML models are always wielded by people or corporations for particular purposes. This is extremely important. "Learning" in machine learning is an aspirational label, not a descriptive one. People who claim otherwise either drank too much of their own Kool-Aid or are…

I mean, at what point will this change? When the AI has to first be trained by being in a robot in the physical world for 10 years learning human concepts before it can start looking at art in the ultimate goal of learning how to draw?

Re: GitHub Copi­lot inves­ti­ga­tion

#902

Post by sofixa »

granzymes wrote: post #318
treis wrote: post #190

Earlier quoted context omitted.

You're not supposed to be able to use dominance in one market (git hosting) to gain dominance in another (AI powered code suggestions).

That might be true, if GiHub were a monopoly. But they are not.

In the same vein that Google are not a monopoly in search, Microsoft are not a monopoly in consumer and business OSes, Chrome is monopoly in browsers, etc.

Just because there is some existing competition which has a few percent marketshare and technically it's not a monopoly doesn't materially change anything besides a pro forma excuse. Which is why Google have been propping up Mozilla, they want the excuse "but technically there's another browser". However for consumers and the market it doesn't matter that technically there's an option that practically nobody uses.

Re: GitHub Copi­lot inves­ti­ga­tion

#903

Post by ghoward »

judge2020 wrote: post #897
cmrdporcupine wrote: post #894

Unfortunately the tone I'm getting from many of these comments makes me feel that people see open source projects as a resource to be mined rather than as a product to be respected. A very entitled attitude ("I really don't want to lose my lovely tool") There seems to be -- on the whole -- little respect for the spirit of the GPL and LGPL and it really is quite a change from, say, 20 years ago, when the 'free softwar…

What do you expect when people grant GitHub an extra license to their repos[0]? 0: https://docs.github.com/en/site-policy/github-terms/github-t...

My code is not on GitHub. If it's there, then someone copied it, and GitHub has no right to claim an extra license to that code.

Re: GitHub Copi­lot inves­ti­ga­tion

#904

Post by nsajko »

judge2020 wrote: post #897
cmrdporcupine wrote: post #894

Unfortunately the tone I'm getting from many of these comments makes me feel that people see open source projects as a resource to be mined rather than as a product to be respected. A very entitled attitude ("I really don't want to lose my lovely tool") There seems to be -- on the whole -- little respect for the spirit of the GPL and LGPL and it really is quite a change from, say, 20 years ago, when the 'free softwar…

What do you expect when people grant GitHub an extra license to their repos[0]? 0: https://docs.github.com/en/site-policy/github-terms/github-t...

This particular licensing term doesn't seem relevant to Copilot.

Re: GitHub Copi­lot inves­ti­ga­tion

#905

Post by tomphoolery »

Does reading software code count as "using software"? I personally don't consider myself subject to a license when I'm reading public code on GitHub. GitHub Copilot and Codex AI seem to be doing nothing more than reading a bunch of source code, not reusing that code to incorporate its functionality into a different product.

Re: GitHub Copi­lot inves­ti­ga­tion

#906

Post by concordDance »

SantalBlush wrote: post #875
concordDance wrote: post #737

Earlier quoted context omitted.

So many comments like this, but none list a tangible harm that's anywhere near the millions of man hours systems like this will save. Heck, basically none (including this one!) list a harm at all!

You could say this about all software piracy. So Microsoft can work to end software copyright instead of trying to corner the market on pirated software and asking forgiveness after the fact.

No, you can't.

Because most software piracy is of saleable software. 20 line snippets are not for sale, mostly because the transaction costs are higher than the snippet value.

Re: GitHub Copi­lot inves­ti­ga­tion

#907

Post by odo1242 »

Vetch wrote: post #642
iepathos wrote: post #554

Earlier quoted context omitted.

Humans do violate copyright if they use copyrighted passages directly in their work and pass it off as their own without any attribution, which is what copilot has been show to sometimes do, though not always. Copilot will sometimes offer chunks of code that can be found verbatim in open source code bases and passes it off to users without attribution. I agree it is ok to learn from copyrighted work and reproduce new…

> which is what copilot has been show to sometimes do In those cases it seems that humans are already copying code without also propagating licenses appropriately. LLMs are more likely to memorize things which occur a lot (and I'd bet rare things that are representative of some conceptual axis). The main examples presented so far, Davis and Carmack, have the property of having been copied a lot. The generative model…

> not just for code

This is quite important, actually, and I don't think enough people realize this. I am a photographer sometimes and it would be really cool if I could share my photos online under a copyright license that forbids their use in training AI.

Re: GitHub Copi­lot inves­ti­ga­tion

#908

Post by makestuff »

I wonder in court if they will rule that this is no different than a human reading open source code to learn how to code. I guess the main difference here is the human would not be able to be used in parallel where Copilot can be used by millions of people at one time.

It will be interesting to see where this goes.

Re: GitHub Copi­lot inves­ti­ga­tion

#909

Post by trention »

concordDance wrote: post #737
Test0129 wrote: post #682

Earlier quoted context omitted.

> It would be sad if someone succeeded in shutting down CoPilot for this kind of copyright stuff. It enables the large scale theft of code. It completely ignores licenses. There are plenty of open source licenses that allow code use with proper attribution yet Copilot doesn't (and probably can't) figure a way to comply with all of them. Copilot, as the article suggests, is a marketing stunt. To me it's more than that…

So many comments like this, but none list a tangible harm that's anywhere near the millions of man hours systems like this will save. Heck, basically none (including this one!) list a harm at all!

Eventually (not in the next 5 years but probably in the next 15) a system like this is going to lower the market salaries of developers (or make some of them outright unemployable). While using their own code as input to achieve that, without ever obtaining actual permission.

Seems like a pretty tangible harm to me.

Re: GitHub Copi­lot inves­ti­ga­tion

#910

Post by an1sotropy »

cmrdporcupine wrote: post #894

Unfortunately the tone I'm getting from many of these comments makes me feel that people see open source projects as a resource to be mined rather than as a product to be respected. A very entitled attitude ("I really don't want to lose my lovely tool") There seems to be -- on the whole -- little respect for the spirit of the GPL and LGPL and it really is quite a change from, say, 20 years ago, when the 'free softwar…

Agreed. Something else apparent from the comments is that people seem to think that some things are copyrighted, and some aren't, and that copilot would be better if it could avoid copyrighted code. Actually, essentially ALL code is copyrighted, and was so the moment it was written, and someone owns that copyright [1]. People only start noticing copyright when the terms of how that copyrighted content is licensed affects them. I think people resent reciprocal licenses like LGPL/GPL because the principle of "share and share alike" that they implement comes with real responsibilities and consequences for the user of the code, while they believe that non-reciprocal licenses (BSD) can be ignored with less serious consequences.

But the show-stopping problem is that copilot is sometimes producing code that is more than fair use of other code that, and is unable to attribute the code or identify how that code is licensed. It is copilot's (Microsoft's) fault that it auto-generates legal minefields, not the person who made an informed decision about licensing their own code.

In spite of the likely downvoting, I'll say that people should be grateful for reciprocal licenses not just because they were and are the foundation of free software (as you point out), but because they shine a light on what it means to license code, and how we are forced to revisit the difference between copyright and licensing when a reciprocal license is violated.

[1] https://en.wikipedia.org/wiki/Berne_Convention

Post reply on HN