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

Copilot sells code other people wrote

twitter.com

Post reply on HN
Print view View original

261–270 of 838 posts

Re: Copilot sells code other people wrote

#261

Post by lin83 »

teakettle42 wrote: post #244
ThereIsNoWorry wrote: post #221

Earlier quoted context omitted.

Ok, then there doesn't exist a single reputable company with a tech division and we're all unethical. Have a nice unethical day.

> Ok, then there doesn't exist a single reputable company with a tech division and we're all unethical. Have a nice unethical day. I’m deeply disturbed that you think this form of plagiarism is universal — I can assure you that is not the case. I work at a FANG currently, and plagiarism is absolutely not tolerated. In fact, plagiarism has been considered a fireable offense at every other company I’ve worked at over m…

> I’m deeply disturbed that you think this form of plagiarism is universal

This thread is an eye opener for me too. Do engineers not get trained on their legal obligations? My company is old and not a tradition tech company but we have been running workshops on the issue for years. Even if they don't, what about their legal teams? Or CI tools to scan for licence violations? Some of the responses here are so naive it's crazy. I hope no one is identifying the companies they work for.

Re: Copilot sells code other people wrote

#262

Post by teakettle42 »

ThereIsNoWorry wrote: post #249
teakettle42 wrote: post #244

Earlier quoted context omitted.

> Ok, then there doesn't exist a single reputable company with a tech division and we're all unethical. Have a nice unethical day. I’m deeply disturbed that you think this form of plagiarism is universal — I can assure you that is not the case. I work at a FANG currently, and plagiarism is absolutely not tolerated. In fact, plagiarism has been considered a fireable offense at every other company I’ve worked at over m…

I'm disturbed you believe regurgitating code snippets is plagiarisation.

It’s literally plagiarism by definition.

Re: Copilot sells code other people wrote

#263

Post by hourago »

hansword wrote: post #203
HumanReadable wrote: post #151

Sorry for the unproductive tone of this comment, but there's something about the attitude of this tweet that really grinds my gears. Any time someone invents something new and incredible, there's always a crowd of negative nancies eager to discredit and explain why the invention is nothing new and a detrement to society. I don't understand why someone would willingly share their code on github where it is publicly av…

If I enter 'Mickey Mouse' into an ML-TTI thing like Craiyon (Dall E mini) do you think I will be able to sell the resulting image on a Tshirt? No, I won't, because Disney has fancy lawyers, the average open source developer hasn't. What you are saying is: Screw little people, let M$ make their money. Either copyright is for everyone, or for no one. I prefer the latter, but this is not the world we live in.

There big difference is that by copying Micky Mouse you are hurting one of the most known and very powerful corporations in the world, by copying code you are just hurting open source projects and individual developers.

It should not be different, or if anything, it should be worse to punish people with less resources. But here we are.

Re: Copilot sells code other people wrote

#264

Post by Closi »

mustyoshi wrote: post #237
meheleventyone wrote: post #219

Earlier quoted context omitted.

One of the (ex?) programmers from Valve managed to get it to spit out parts of the Source engine verbatim. He posted a Twitter thread yesterday I believe.

Does that prove it ignores licenses or does that imply the source engine exists verbatim (minus licenses) multiple times on Github?

It would prove that it doesn't honour all licences - just because the source code exists on Github without a licence doesn't automatically grant a licence to Copilot from a legal perspective.

Re: Copilot sells code other people wrote

#265

Post by Aeolun »

tagyro wrote: post #162
coldtea wrote: post #86

> Hector Martin: If you use Copilot, you are basically playing Russian Roulette that the random mashup of existing, copyrighted, hegerogenously licensed code that you get out of it qualifies as an original work, mostly by chance. Or that nobody will ever sue you otherwise. Well, that's already the case with Stack Overflow copypasta enterprise code. If anything, use of Copilot would be an improvement...

Do people really copy/paste from StackOverflow? I feel this is more a meme, rather than reality. I do check StackOverflow, but never have I took an answer verbatim. I try to see if it's the same problem and what was the approach in deconstructing it, which I find more useful in the long run.

This has more to do with the code never being immediately copy-pasteable, not so much my reluctance to copy-paste from SO for licensing reasons.

Re: Copilot sells code other people wrote

#266

Post by ThePhysicist »

HumanReadable wrote: post #151

Sorry for the unproductive tone of this comment, but there's something about the attitude of this tweet that really grinds my gears. Any time someone invents something new and incredible, there's always a crowd of negative nancies eager to discredit and explain why the invention is nothing new and a detrement to society. I don't understand why someone would willingly share their code on github where it is publicly av…

I mean I'm not an expert but it's a valid point as people share code under a given license, and as far as I'm aware Copilot does not make this knowledge available. Nothing to do with the fact that Copilot is an amazing technological achievement.

If I, as a human, go to a public repository on Github and copy/paste a non-trivial 200 line code snippet into my proprietary code base I have to abide by the license of that original code, even if I slightly modify it. I don't see how this cannot be true for Copilot. I'm sure the legal folks at Github have thought of a response though, you could e.g. argue that the snippets produced by Copilot are not affected by the copyright of the original author as they do not reach the required treshold of originality. Seems rather shaky for me though.

Re: Copilot sells code other people wrote

#267

Post by avereveard »

swader999 wrote: post #156
bloat wrote: post #82

Earlier quoted context omitted.

This is interesting. Is the internet completely cut off? Do they have internal libraries of documentation for third party stuff they are using (paper? digital?) Do you have any example institutions, or what domain they are working in? Thanks.

I think it would be for super secure military coding. But business domains? Hardly ever.

The issue doesn't solely rest in copyright

A concern, which I think is legit, is that it is quite easy for someone with a strong presence in search, web advertising, analytics and mobile to puzzle together what a company is investing in based on the aggregated research and web access from known locations

Re: Copilot sells code other people wrote

#268

Post by micromacrofoot »

mustyoshi wrote: post #237
meheleventyone wrote: post #219

Earlier quoted context omitted.

One of the (ex?) programmers from Valve managed to get it to spit out parts of the Source engine verbatim. He posted a Twitter thread yesterday I believe.

Does that prove it ignores licenses or does that imply the source engine exists verbatim (minus licenses) multiple times on Github?

just because someone else ignored the license doesn’t mean github is free to blindly vacuum that up

Re: Copilot sells code other people wrote

#269

Post by ThereIsNoWorry »

teakettle42 wrote: post #262
ThereIsNoWorry wrote: post #249

Earlier quoted context omitted.

I'm disturbed you believe regurgitating code snippets is plagiarisation.

It’s literally plagiarism by definition.

https://stackoverflow.blog/2021/12/30/how-often-do-people-ac...

I feel like you're arguing in bad faith. So, whatever.

Re: Copilot sells code other people wrote

#270

Post by kaetemi »

npteljes wrote: post #128
lakomen wrote: post #41

I don't understand what's going on there. I don't use github. Can someone explain what the author means? Edit: in detail

GitHub Copilot is a paid feature, but that's a red herring in this discussion - people are free to monetize free software, neither or the major licenses forbid this. GitGub Copilot is an advanced autocomplete / code generation system, based on a machine learning model. The code used for training the model is taken from projects hosted on GitHub. These projects were published under different licenses. The main questio…

Those instances where people demonstrate verbatim copies, are mostly either well known snippets which have been copied a million times already, or obvious completions of a partial verbatim piece of the supposedly copied code that any coder could extrapolate.
Post reply on HN