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

381–390 of 838 posts

Re: Copilot sells code other people wrote

#381

Post by aaaaaaaaata »

tremon wrote: post #341
jimnotgym wrote: post #285

Earlier quoted context omitted.

Just because there has not been a test case yet does not make it illegal! If MS think it is fair use then they are free to go ahead. Business is all about recognising and assesing risks like this.

And even if there had been a legal test case, that does not make it moral! If people think this is socially wrong then they're free to argue their case. Business is all about ignoring ethical quandaries if it gives them an edge. "Microsoft does it, therefore it must be right" does not a sound argument make.

> Business is all about ignoring ethical quandaries

No, businesses are — not business. Not necessarily...

Re: Copilot sells code other people wrote

#382

Post by drexlspivey »

meheleventyone wrote: post #158
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…

They own their code and it either has a license for use or is implicitly rights retained if not. If Copilot regurgitates their code, from a project that is public but with a non-permissive license they are having their IP rights violated so are totally correct in being unhappy about that. Just because you’ve made something cool doesn’t give you the right to harm others in the process. If MS or OpenAI don’t think this…

Owning code snippets sounds ridiculous to me, like can I own this snippet?

    def average(*numbers):
        return sum(numbers)/len(numbers)
if not is it because it is too small? what’s the minimum line number that ownership kicks into? what if I change the function name and the variable names?

Re: Copilot sells code other people wrote

#383

Post by anonymoushn »

wolframhempel wrote: post #290
dmortin wrote: post #110

Earlier quoted context omitted.

Did the scan find the process if they changed the variable names, for example? Or is that considered a differing snippet then?

This is exactly where it gets murky. We had the usual 1-4 line snippets. We went the extra mile to change them, rewriting them from scratch, partially with different implementations. Did we need to do that? Would it have been enough to just change a variable name or some spacing or similar? I don't think there's a clear standard. The music industry has struggled with this for a long time. When is a song derivative, w…

That sounds rough. Here's an 8-line snippet, please make sure you don't infringe my copyright:

    p = mmap(
        null,
        size,
        PROT_READ | PROT_WRITE,
        MAP_PRIVATE | MAP_ANONYMOUS,
        -1,
        0,
    );

Re: Copilot sells code other people wrote

#384

Post by lupire »

gumby wrote: post #373
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…

People get paid to write code having learned from writing code for others and from reading code others wrote. In this regard I dont see why github copilot is any different.

People don't memorize chunks of code and copy it.

Re: Copilot sells code other people wrote

#385

Post by causi »

lupire wrote: post #372
causi wrote: post #291

Earlier quoted context omitted.

Unfortunately the way IP law works, at least in the US, is that you can use essentially whatever you want as training data and it's up to the user to make sure none of the generated code violates licensing agreements.

Did you just make that up? Github is distributing the copied code to users.

Did you just make that up?

Unfortunately not. It's really stupid.

Re: Copilot sells code other people wrote

#386

Post by nirvdrum »

jimnotgym wrote: post #295
meheleventyone wrote: post #223

Earlier quoted context omitted.

Is this fair use? I don't think that's been established yet. And if it is why didn't MS and OpenAI train it on their private code repositories? Fair use for thee not for me isn't very in keeping with the spirit of that claim.

I sometimes read people's open source code on github and use the ideas from that to develop my own ideas. In fact sometimes I copy and paste short passages and then rework them. I also employ a team of people who may do the same. Is that fair use, yes of course it is. Is co-pilot automating that fair use, I would say so.

Many people would claim what you're doing is a derivative work. I'm not sure the "of course it is" is very clear-cut (at least in the US). I've worked at big companies that have lawyers that care very much about this topic and what you're describing is prohibited. But, maybe it's different if you're not distributing your source.

Re: Copilot sells code other people wrote

#387

Post by kop316 »

jillesvangurp wrote: post #358
meheleventyone wrote: post #158

Earlier quoted context omitted.

They own their code and it either has a license for use or is implicitly rights retained if not. If Copilot regurgitates their code, from a project that is public but with a non-permissive license they are having their IP rights violated so are totally correct in being unhappy about that. Just because you’ve made something cool doesn’t give you the right to harm others in the process. If MS or OpenAI don’t think this…

I'm sure the MS lawyers thought long and hard about this and are patiently awaiting any actual lawsuits with confidence in their position. It would be very hard to prove ownership of any snippets. To the point where you can argue that it is just fair use and to the point where companies would think long and hard before committing any resources to fighting MS on this in court at great expense. I don't think that will…

> It would be very hard to prove ownership of any snippets. To the point where you can argue that it is just fair use and to the point where companies would think long and hard before committing any resources to fighting MS on this in court at great expense.

I would like to point you to this: https://twitter.com/mitsuhiko/status/1410886329924194309 HN Comments at the time: https://news.ycombinator.com/item?id=27710287

Re: Copilot sells code other people wrote

#388

Post by Havoc »

Yes, though in a way so does stackoverflow & friends. Large chunk of dev ecosystem is copy paste and I don't think this is inherently problematic. It is always a case of standing on the shoulders of giants.

Its more of a licensing issue to me. As far as I can tell it was train on a blend of licenses which to me makes it inherently non-compliant. At least some of it is going to be copyleft and find its way into closed source.

Re: Copilot sells code other people wrote

#389

Post by ay »

tiborsaas wrote: post #331
moffkalast wrote: post #298

Earlier quoted context omitted.

> it's not repeating non-trivial work 1:1 But that's basically all copilot does? It's just a fancy compression system with a search function.

No, it customizes the snippets to your context, the code is synthesized and not looked up in a db like a web search engine.

I tried it for the first time today, so treat this with a grain of salt.

https://twitter.com/ayourtch/status/1539928018138931200 is my experiment. The code in question has a very specific format - it’s C with a lot macro sauce. I described the intent in the comment and pasted the includes lines. Then I started the #define of a unique looking token, and it added the lines with the correct boilerplate. What you see in gray is more boilerplate that it suggests when prompted.

I would dare to assert that “xxxayourtchtestxxx” is not going to be in anyone else’s code than mine.

So you can see the example of copilot generating completely new code.

Not saying it’s 100% of what it does - but this side looks very useful.

I also did a test with Rust: described a function canonicalizing MAC address, and then when it saw ![test] prompts, it started to make very passable unit tests for the function which was not even written yet - it was only the comment of what it would do.

Also a massively useful lever to have, if it can do so consistently.

My attempts to make it generate a bug-free canonicalization function didn’t work - but it was interesting to see it try different approaches based on the existing test code (and no, they didn’t always satisfy the tests, unlike one would expect :)

So this angle is “pair programming with a creative novice”, which also can be useful - it can give ideas to explore that you didn’t think of.

Of course this was all fairly trivial code, I do not know yet how it will behave in a more tricky situation.

Re: Copilot sells code other people wrote

#390

Post by ParetoOptimal »

olalonde wrote: post #345
ParetoOptimal wrote: post #337

Earlier quoted context omitted.

You can be both a professional software developer and a caring human that considers ethics and exercises empathy. It's easy to convince oneself they can only be a professional developer to escape ethical responsibilities which require significant time and energy.

Of course you can and you should. But going from the Twitter bio and personal website, it doesn't appear to be the case here. They're an activist who lives from soliciting donations and selling 30$ videos on how to be anti-racist (like, literally).

> They're an activist who lives from soliciting donations

Assuming this is the case, are you concluding they don't have time to be a real software developer?

Maybe they were a professional developer, but now are an activist 50% of the time.

> selling 30$ videos on how to be anti-racist

Is the indictment here that they are a capitalist?

My basic point is you seem to really want to dismiss their views you don't like by arguing they aren't credible rather than attacking their ideas.

Post reply on HN