> Tell me you don’t use Copilot without telling me about it.

You don't accept arguments against the use of Copilot from people unless they... use it?

That's a nifty way to ignore any and all criticism of Copilot, or indeed any discussion about any ethical issue ever.

They should at least try to understand how it’s actually used, not imagining how it’s simply used to steal their largely replaceable code.

It doesn't matter how it's used. Do you think Microsoft would be happy with someone training a model on Windows source code, as long as they didn't use it to reproduce the code?

If Microsoft were confident Copilot doesn't produce infringing code, they would have included the Windows and Office codebases in the training data. I wonder what will come out of discovery

You think MS‘ code quality is high enough to train an AI on?

Do you think they audited every open source code base that was used in training for quality?

“Their largely replaceable code”

Smells like: “ I stole this lousy apple that wasn’t any good” Then why did you steal it?

Put your money where your mouth is, Microsoft, train copilot on your own code!!!

Don’t wanna train it with windows 11 code? Prefer to hijack others projects and use their for your needs and then pretend thst insulting others and calling their code worthless will get you off the hook????

Backfire

> Smells like: “ I stole this lousy apple that wasn’t any good” Then why did you steal it?

The lousy code trained copilot in what a switch statement looks like so it can autocomplete mine for me

>> Tell me you don’t use Copilot without telling me about it.

> You don't accept arguments against the use of Copilot from people unless they... use it?

> That's a nifty way to ignore any and all criticism of Copilot, or indeed any discussion about any ethical issue ever.

"I only listen to people who agree with me, but to make that sound legitimate, I have a somewhat indirect way of saying so."

I believe the argument is that you shouldn't accept arguments against the use of copilot from people unless they have tried to use it. In a realistic context. That seems reasonable to me. It's the bare minimum to make an informed opinion. I think the wording was perhaps poor, but I think your interpretation is a little reductive/disingenuous.

I don't see how using copilot make the copyright question any less serious.

Because it's useful, then it's not a problem?

Well, it's also useful to send our non recyclable trash to 3rd world countries and every 1st World country should try it. It will definitely make the consequences less serious if everyone does it.

Not apples to apples but I guess you get the picture.

Because its a net good to the world its not a problem. It the benefit is orders of magnitude greater than the harm then its good.

It’s a net harm for the programmers whose code is being willfully plagiarized.

It’s a net boon for Microsoft in their efforts to rule the world.

It’s a net loss for society and ethics.

Open up copilot code, Microsoft, if you are so sure that everyone must wear transparent underwear let’s see you wearing some. Train copilot on windows 11 code. It’s not public domain.

Truth matters. Lies matter

Expand on the unethical part. So people published code that could be referenced and copied on GitHub. There was no ethical problem, the world, society were happy.

Github make a convenient way to search and contextualise this publicly available code and paste it into your code (adjusting local scope, format, language along the way). Suddenly we have crossed an ethical line!?

Which ethical line? Are you pretending people never copy and pasted open source code before copilot? Are you pretending open source code never copy and pasted other open source code? That we were in an ethically pure world until copilot came along?

> So people published code that could be referenced and copied on GitHub. There was no ethical problem, the world, society were happy.

This code has different licenses. You can't just copy code randomly without checking license first.

Copilot serves it stripped of the license to unaware users. Even if copilot user wants only to reuse code licensed in a way that allows it copilot will serve him code from restrictive licenses without him being aware.

You can just copy and paste code without checking the license. People do it all the time.

GitHub doesn’t force you to accept the license in the repository before showing you the code.

> It’s a net harm for the programmers whose code is being willfully plagiarized.

What's the harm, specifically?

Say it copies that snippet of workflow scheduling code I made at work yesterday or the greasemonkey script I made in my own time.

How is my life worse?

I believe the argument being made is that in _actual, real-world_ use of copilot, no copyright infringement happens. In order to make an informed decision as to whether you agree with that, you can try copilot to reach an informed conclusion. There is no cost to try it. Unlike your example, where "trying" has an immediate cost -- which is why that example doesn't make too much sense here.

Yep know it's much more clear, thanks for clarifying.

See your sister comment's child for my reply.

As someone who has used copilot since the early beta days, what I think people are saying is that nobody uses Copilot to generate full functions like this. It's more of an intelligent auto complete. It's fantastic for repetitive autocomplete where certain things have to be changed, and for quickly getting out boilerplate code. You can put a list in a comment along with a format and generate data structures quickly and easily. You can solve small problems quickly, allowing you to focus on the bigger picture.

It's sort of like a power tool — sure, you could use a screwdriver, but a drill with a screwdriver attachment will be quicker. Hammers are good, nail guns are quicker. You'd never expect someone to use a drill with sd attachment if they'd never used a screwdriver before.

There are for sure things to be improved, such as the recent post on how you could put in a very specific seed and get out a specific function that it shouldn't. The answer here isn't to shut down the project, at a net loss for everyone, but to find ways to improve it.

As others have said, with Copilot gone and the new demand created, the vacuum will bring in community projects that will happily scrape every public repository they can get their hands on.

Now I understand what you mean, still pretty crappy because those minor auto complete strings only exist because Microsoft used code without permission and/or without crediting the original owners and/or breaking licenses from the original corpus.

I use copilot everyday, I love it, but it still leaves a bad taste in my mouth knowing that people out there worked really hard on their code and harder on building OSS licences just for Microsoft to throw all that out of the window.

Feels like licenses don't matter anymore. My own code doesn't matter much, but it's about principals dude, licenses are there and they should be respected, if not, then it's just anarchy, and we all know anarchy only works in very specific scenarios, Microsoft is not apart of any of those scenarios.

> I believe the argument is that you shouldn't accept arguments against the use of copilot from people unless they have tried to use it.

I disagree, and this does not hold up generally: We can, and should, argue things we have not tried or experienced, like heroin and murder. What makes it so that this has to be tried?

> It's the bare minimum to make an informed opinion.

Only if the usefulness is what is in question. But it is not.

Correct, it doesn't hold up generally. But it doesn't need to. It holds up here. We do not try things when there is exceptional risk or cost in the trying. Here there is no cost to trying, so it does not make sense not to try.

I believe the argument being made is that in _actual, real-world_ use of copilot, no copyright infringement happens. So it's not just about usefulness.

> I believe the argument being made is that in _actual, real-world_ use of copilot, no copyright infringement happens. So it's not just about usefulness.

How would you know though? The burden of proof is on Copilot. Especially now that it has been shown to spit out copyrighted code.

You’re right, trying Copilot is equivalent to committing murder.

/s

> We can, and should, argue things we have not tried or experienced, like heroin and murder. What makes it so that this has to be tried?

It's not that you absolutely have to have experience with something, but you'd be foolish to discount the input of people who do. In debates about drug policy I try to be polite to people with zero first hand experience, but their contributions are rarely of interest. Murder is a bit more abstract insofar as anyone who has fully experienced it by definition didn't survive to testify, but I give a lot more weight to the views of people that have first-hand knowledge of violence and crime.

It's not that you shouldn't weigh in on a topic without first hand experience, but that it's a good idea to specify the scope of your understanding, or frame uncertainties as open questions rather than assumptions.

On a different website I argued with a microsoft employee who said that copilot is great and so on and would not discuss unless I tried it.

I tried telling that it requires a credit card number to try it but he didn't believe me… I guess the thought that non-microsoft employees have to pay for microsoft stuff didn't occur.

That isn't sufficient to get you off the hook. Copyright covers derivative work, not just code that's reproduced verbatim.

Derivative work requires a major part of the original, before it’s considered by copyright.

A 3 line boilerplate is neither novel nor a major part of the original.

I think if copilot is heavily restricted to three line code samples, perhaps I could agree with this.

The example cited by the OP is not a three line code sample - if you've ever done matrix coding, you know that sparse matrix operations are not simple.

Sure you can reduce it to a function call, but then you have library usage instead of code theft.

I think actually perhaps this is a way copilot could ethically move forward - instead of lifting code verbatim if it merely suggested libraries and approaches "here is an example of sparse matrix filtering and some libraries which do it", that would be both useful and ethical, presuming it does not obscure the license.

...But the example cited by the OP isn't how anyone actually uses copilot...

From where I sit, the complainant has found an extremely convoluted (and buggy) way to copy-paste their own code and is very upset about it. By similar logic, we should restrict the use of ctrl-c and ctrl-v, because they allow very simple infringement of open source licenses. Find a sparse matrix multiplication library which uses the copied code without attribution and you can take them to court; the law is already sufficient for this.

"Derivative work" is a very specific thing, and it's contrasted with "transformative work" in a way that matters a lot, and fair use intersects heavily with both.

Even when it comes to stuff that seems reaaaaally close to pure derivative: Googling "How long does it take to boil water?" => "If you're boiling water on the stovetop, in a standard sized saucepan, then it takes around 10 minutes for the correct temp of boiling water to be reached. In a kettle, the boiling point is reached in half this time."

That's a verbatim snippet pulled directly from https://unocasa.com/blogs/tips/how-long-to-boil-water, and yet Google exists and continues to do stuff like this under the fair use doctrine despite massive efforts to attack/monetize their service. [To be fair, Google does link results, which probably insulates them because it's less hurtful to the commercial interests of the source; that said, with open source there generally are no commercial interests to hurt (open source attribution will be a tough sell as an actual commercial interest), and that's specifically called out in the law as a factor]

Copilot is even less explicitly at risk IMO, in that it never even stores the text, nor can it reliably retrieve it. I have no idea what makes anyone think it should be more vulnerable than Google.

From the copyright.gov page on fair use (https://www.copyright.gov/fair-use/, worth reading in detail for anyone who cares about this stuff, also has links to a monumental number of cases with shockingly intelligible summaries): "Additionally, “transformative” uses are more likely to be considered fair. Transformative uses are those that add something new, with a further purpose or different character, and do not substitute for the original use of the work."

Copilot without any shadow of a doubt does add something new, with a further purpose, and does not substitute for the original use of any codebase on Github (it can't create any of the codebases in full, without manual guidance so extreme that you'd have to be using the actual original codebase as a reference, so it clearly cannot substitute for a single one of them, and that's what a lawyer will argue, likely successfully).

In the Google vs. Oracle case (see https://www.copyright.gov/fair-use/summaries/google-llc-orac...), a big piece of the fair use finding was that "its value in significant part derives from the value that those who do not hold copyrights, namely, computer programmers, invest of their own time and effort to learn." and "further[s] the development of � computer programs”. It's hard to see where Copilot wouldn't fall into that category, as well, and that's precedent on (multiple) appeal.

By my reading this should be a slam-dunk fair use ruling, unless precedent gets really upended, and Butterick is wasting a ton of time and effort for absolutely zero potential gain other than some bragging rights, but to each his own...I guess we all have to grind our axes from time to time.

"Derivative work" is a very specific thing, and it's contrasted with "transformative work" in a way that matters a lot, and fair use intersects heavily with both.

Even when it comes to stuff that seems reaaaaally close to pure derivative: Googling "How long does it take to boil water?" => "If you're boiling water on the stovetop, in a standard sized saucepan, then it takes around 10 minutes for the correct temp of boiling water to be reached. In a kettle, the boiling point is reached in half this time."

That's a verbatim snippet pulled directly from https://unocasa.com/blogs/tips/how-long-to-boil-water, and yet Google exists and continues to do stuff like this under the fair use doctrine despite massive efforts to attack/monetize their service. [To be fair, Google does link results, which probably insulates them because it's less hurtful to the commercial interests of the source; that said, with open source there generally are no commercial interests to hurt (open source attribution will be a tough sell as an actual commercial interest), and that's specifically called out in the law as a factor]

Copilot is even less explicitly at risk IMO, in that it never even stores the text, nor can it reliably retrieve it. I have no idea what makes anyone think it should be more vulnerable than Google.

From the copyright.gov page on fair use (https://www.copyright.gov/fair-use/, worth reading in detail for anyone who cares about this stuff, also has links to a monumental number of cases with shockingly intelligible summaries): "Additionally, “transformative” uses are more likely to be considered fair. Transformative uses are those that add something new, with a further purpose or different character, and do not substitute for the original use of the work."

Copilot without any shadow of a doubt does add something new, with a further purpose, and does not substitute for the original use of any codebase on Github (it can't create any of the codebases in full, without manual guidance so extreme that you'd have to be using the actual original codebase as a reference, so it clearly cannot substitute for a single one of them, and that's what a lawyer will argue, likely successfully).

In the Google vs. Oracle case (see https://www.copyright.gov/fair-use/summaries/google-llc-orac...), a big piece of the fair use finding was that "its value in significant part derives from the value that those who do not hold copyrights, namely, computer programmers, invest of their own time and effort to learn." and "further[s] the development of computer programs”. It's hard to see where Copilot wouldn't fall into that category, as well, and that's precedent on (multiple) appeal.

By my reading this should be a slam-dunk fair use ruling, unless precedent gets really upended, and Butterick is wasting a ton of time and effort for absolutely zero potential gain other than some bragging rights, but to each his own...I guess we all have to grind our axes from time to time.

> Copilot is even less explicitly at risk IMO, in that it never even stores the text, nor can it reliably retrieve it. I have no idea what makes anyone think it should be more vulnerable than Google.

That doesn’t matter, IMHO. Once Copilot manages to copy a work, a copy has been created and copyright has been violated. If this occurence is reasonably likely, then Copilot is wittingly assisting in violating copyright.

How likely that situation is is disputable. People have cherry picked some cases where by specifying a very particular comment you can get it to (unreliably) reproduce well-known pieces of code, sure. But that is neither the intended use of the product nor the actual way that a single person using the product really employs it, and that really, really matters. Judges are not automatons, and the fact that when and if this goes to court the developers will be able to get up and honestly say "this is a tool developed to help developers create new code that is not merely recreating the functionality of the code we trained the tool on, and all our users use it to create entirely new things" is going to matter when they argue that it is fair use/transformative work.

I do completely understand that a lot of people disagree with me morally, and think that extracting insights from scraping the public web should be illegal. You're free to have that opinion, but I'd recommend you start lobbying your congressman to change the law, because though I'm not a lawyer I hang out with a few who do copyright stuff, and I don't think the law as it stands is on your side. That said, who can say, maybe this will end up bubbling up through many layers of appeals and end up at the Supreme Court someday, this stuff is all certainly wildly outside the bounds of what anyone writing the copyright laws was thinking about back in the 70s (which I think was the latest significant iteration?) so it's fair to say it's a complete gray area.

>But that is neither the intended use of the product nor the actual way that a single person using the product really employs it, and that really, really matters

And yet, someone has done so, and found that to be the case. Therefore, it falls under "normal use".

So make up your mind. Is software engineering only about assembling working programs, or assembling working programs + navigating license minutiae?

One of these, Copilot has a place in. The other it does not.

You may hope that a judge considers a really weird and cherry-picked weird edge case to be "normal use", but opposing counsel will argue the opposite, and have a lot of evidence if they actually look at real use.

"I can use a gun to shoot someone" doesn't make guns illegal, even if people do so with some regularity. "I shot someone to make the point that guns can kill people" is worth even less in the eye of the law, and that is literally what you're pointing to here.

And I don't know what you're even getting at with the "assembling working programs" vs "navigating license minutae" stuff. Copilot helps the former, and you're apparently trying to argue that it should be banned because some people are feeling angsty about the latter?

Google then handily provides you A LINK to the original ATTRIBUTED SOURCE.

Yes, I mentioned that. And I think it would be great if GitHub added some sort of scan to see if it was accidentally returning some verbatim code and gave a link to the repo it came from, but the situation with Google was much more clearly harmful: by scraping and presenting verbatim quotes, they are directly siphoning web traffic away from the sites they scrape from, and many of those sites make money based on page views. That is direct economic harm, and the actual impact was significant for many companies, not merely theoretical.

With open source code, the harm is much less tangible, since negligibly few open source projects make money from people going to their GitHub pages because they're searching for code snippets (not zero, but almost). My guess is that an honest quantification would put the lost revenue due to Copilot's existence in the tens to maaaaaybe hundreds of dollars. Courts look at that type of thing, which is why I don't think this will end up being an issue, at least in the US. Europe is wild, who knows what they'll do there, and that's where activists on this topic should most wisely apply pressure, you can always convince someone in government there to throw a spear at a BigCo. You won't take them down, but you may get them to negotiate, and I don't even necessarily think that's a bad thing.

That said, even in the US, if enough people make noise then things could change, so I encourage you to speak to your congressperson (I will be as well, but arguing the other side, because I really do think this is fair use and I'd like to see it enshrined as such explicitly, because this fight is going to be extremely common over the next few decades).

Ok so you don't care about the second part of the argument, you don't care about the danger of violating copyright.. The first question though is if training on large bodies of OSS is fair use. Are you saying you don't care about the copyright of your own work either? Or you don't care because you don't publish your work OSS?

I’m sure I’ve done more to OSS than 95% of commenters here. I publish my code under MIT when possible (and WTFPL for smaller projects), and yes, please train on my work or split out my functions verbatim, they are far less valuable than some people seem to believe. I don’t even care about the attribution part of MIT, it’s simply a nice to have when decent people use my code.

This isn’t just about OSS. If I have a public repo on GitHub without a license, you can’t rewrite that code and put it in your project. I own the copyright. The issue is that Copilot will still launder it into your project for you.

Oefrha just said that his, and probably yours, code is not worth that much if splitted into functions and commentary blocks. The value of software comes from its whole purpose, not your clever email-validation-regex you are so proud of.

> not your clever email-validation-regex you are so proud of

Really? I challenge you to write a correct email validation regex.

>code is not worth that much if splitted into functions

Some algorithms in scientific computing require lots of effort to implement as nice, reusable, performant function. Those functions often more important than whatever the whole is doing because it's what most other people will be interested in using.

Then why use OSS licenses in the first place?

Let's just copy each other's code without attribution.

I am totally for this! Will bring eceryone forward in general without costing the society as much as this copyright bs.

Either that or people will stop posting their code publicly.

How many lines of code was the Oracle v Google case? 9?

Just so happens I think the case is bullshit. You would have a point if I supported Oracle.

Your lack of support of them doesn't change the facts.

Says you.

If it’s all just disposable code, WHY ISNT MICROSOFT TRAINING COPILOT ON WINDOWS AND OFFICE CODE?

Because that code isn't publicly available, just like every private repo on GitHub, and any other company's/person's proprietary/private code?

Again, if I put code in a public repo on GitHub and don’t include a license that allows it, you cannot copy that code. It doesn’t matter that you’re able to access it.

Cannot is a strong term. Telling me there may be consequences would be more accurate.

I can claim fair use and short of an injunction there's nothing you can do to stop me.

Claiming fair use is not a fix all card you can play like that, there are a lot of nuences for that to hold on trial.

I'm allowed to read it and learn from it. To distill the core concepts and recreate it.

You as a human being, not you as a for profit company building a for profit product.

Maybe I’m the worst programmer in the world. It doesn’t matter. My code is still my code, and if I don’t explicitly license it such that you can copy it, you’re not allowed to.

Wait... Shouldn't you use a different license if you don't care about attribution?

Creative commons maybe.

Just because you, as someone who self proclaimed to have done more OSS than 95% of the commenters here, does not know how to use OSS licenses, doesn't mean that the copyright question being discussed here is a non issue.

The issue is that you don't care about what the licenses in your code mean:

> I publish my code under MIT when possible (...) please train on my work or split out my functions verbatim (...) I don’t even care about the attribution part of MIT

I choose MIT because it’s a widely used permissive license most people are confident about using, not because I will personally pursue every clause in it. I will not take action against anyone using my MIT-licensed code without attribution.

Having been a member of a very high profile permissively licensed project and having started a few relatively popular ones of my own, I’d say I don’t need to take licensing advice or be called “does not know how to use OSS licenses” from someone who laughably advises using Creative Commons, when Creative Commons itself advises against using CC licenses for source code, except CC0, which is entirely different from CC.

There is MIT-0, a no-attribution variant of the same license.

My good, sorry to move away from the topic, but do you realise how childish you look with your comments?

I was considering not replying, but here goes nothing....

> not because I will personally pursue every clause in it.

Then you are out of the game, because all clauses should be respected, or else you are committing something very close to an illegality when you violate said clauses. If you don't want a specific clause, consult a lawyer and remove it, or use another license.

> Creative Commons itself advises against using CC licenses for source code

So before your didn't care about respecting clauses and now you do?

I could argue: there are some bits of text in CC that make it not a good license for code, but I don't care because I don't respect those clauses. I'm not going to make that argument, because it doesn't make any sense. You either use a license and respect it or you don't.

Just because you don't care does not mean others should not care, even if they are less valuable coders than you.

I mostly contribute by finding issues and reporting them, does that make me less of a OSS contributor than you?

And yet I do care if my private project is used by behemoth like Microsoft without my consent, even if it's only poorly written fizzbuzz. Why? because if I wanted to share it I would publish.

I was answering the questions “Are you saying you don't care about the copyright of your own work either? Or you don't care because you don't publish your work OSS?”

Feel free to care.

I personally don't care about the copyright of my own work when it comes to code. I publish the majority of my work anonymously with a clause that just says something along the lines of "feel free to use some or all of this without any attribution." Denying others from using my code feels like it goes against the essence of OSS itself. It's like if I found the solution to some complicated math problem but didn't let anyone else see it because it's "mine."

You are both a scholar, and a gentleperson. Thank you for treating the field like it is instead of some social value extraction platform.