I can't decide if people just hate Microsoft enough that a future where you must pay to include an iseven function in your code is a price worth paying to give them a bloody nose, or there are just a large contingent of users making millions off their GPL code who are put out.

More seriously yes, copilot damages copyright (or is perceived to) and that is a good outcome irrespective of the actor. I will never see eye to eye with people defending the existing legal framework.

Suppose a commercial software company just took GPL-licensed software and openly incorporated into their own code and then sold. Would that "damage" copyright also? Remember, there's no legal principle that says "if we catch violating copyright, your stuff is now free". The copyright holder can sue for damages or to stop distribution and that's it.

People like Larry Ellison of Oracle have claimed they just steal GPL'd stuff 'cause it there. But Oracle defends it's copyrighted code very aggressively. Oppositely, the GPL is intended to allow more open access than public domain in a time where commercial companies want to take anything they can get.

My code is 100% in Github Pilot, is there any way to publicly say that I'm against the lawsuit even if they pretend to represent me?

1) If Github Copilot is a free software liscensed under GPL, I'm all for it. Microsoft is using other's collective hard labour to benefit itself.

2) Its Microsoft. The king of dark patterns, monopolization, and the enemy of software freedom. They can't have their cake and eat it too.

Is it patent trolling when you are defending your future labor from being made obsolete by megacorps and signularitarians using your past labor without permission?

No, it's justice.

The entire response to this suit on this site is mind-blowing to me. Everyone is up in arms that someone trained an AI model that could potentially spit out tiny, twisted fragments of public, open-source code. This response is nothing but selfish behavior that runs counter to the core principles of open source development and the free software movement.

But training an AI model on media (code or otherwise) is not copyright infringement, so the license is irrelevant.

It's selfish to pretend otherwise and to try to assert a copyright right that doesn't exist, for the purpose of impeding progress in a field that benefits us all.

>the right to exclude others from making certain uses of the work: copying it, making a derivative work based on that work, distributing copies of the work to the public, and publicly performing or displaying the work.

So why would "training" "AI" on code with the intention of emitting derived works not be copyright infringement exactly?

This product is transforming copyrighted code into something that's intended to be used or sold in other works. The snippets it emits are directly derived from copyrighted code.

The most common argument against this is that humans also learn from copyrighted material. My argument against this is that CoPilot is not a human and should not be assumed to inherit rules intended for humans.

>in a field that benefits us all

As it stands currently CoPilot is proprietary and does not benefit anyone except for MicroSoft. If CoPilot was released under a FOSS license it would actually benefit us all. Most of the people against CoPilot are not against AI, but rather a proprietary AI product transforming FOSS work into other potentially proprietary works with the intention of profiting off of the completion service and hoarding the code that powers it.

The fact that you write something, doesn't automatically make that thing true.

Some uses of AI were ruled not to be infringement. This is a different case which requires a different ruling.

Well, maybe. But even if we assume that this is true, when anyone later uses the AI to reproduce a copy of the code, a copy has been made and copyright has been infringed.

If I need a code to loop over 10 lines, I'll code a for loop the same way regardless of what I'm developing.

Define for me, at what point of complexity, does code gets Copyrighted?

The things copilot is outputting is literally small chunks of code that needs a lot of cleanup afterwards. Is not like I type "Build twitter for me" and BAM, I got a working clone of twitter.

copilot is outputting literally small chunks of code that needs a lot of cleanup afterwards

If you start from copyrighted chunks, then clean them, you're still violating people's copyright. Multi-million dollar lawsuits have been fought over people using small samples from other people's music, cleaning them, and releasing them as parts of their own song.

I don't care if people make money using GPL'd code, but I do care if they take the code and strip the license so they can use it in non-free projects.

I just hope Microsoft AutoPlagiarist is not the Final Solution to Free Software they have been seeking since before the millennium's turn.

Seems to me this discussion is likely to pivot on a fulcrum located between "old enough to remember Microsoft before Bill Gates began spending his ill-gotten gains on philanthropy" and "young enough to see Microsoft primarily as the Xbox people".

What terrible outcome will we see from a lack of copyright law?

Please do name one industry, niche or platform where copyright does actually prevent this from happening in any meaningful way today.

More generally, what is left to protect any creative work besides guarding physical access? Why would any company make any movie or tv show if it could be copied and redistributed by others endlessly the moment it gets shown once?

There have been creative endeavous before copyright and there would be creative endeavours after copyright. Perhaps even more since people are free to remix and share without restrictions.

This is mostly because the means to copy requires little effort compared to the act of creating. So, there is no incentive to create because you wouldn't make a living out of it. Imagine spending two years writing a book and someone buys one and copy it to sell at 25%. He can make a profit at a lower threshold than you, so you as a creator cannot compete.

That's true even in USA (with strong copyrights), apart from for top .1%

To be fair, from my experience most countries don't have much of a movie scene even with copyright and instead mostly import hollywood stuff.

> This is mostly because the means to copy requires little effort compared to the act of creating. So, there is no incentive to create because you wouldn't make a living out of it. Imagine spending two years writing a book and someone buys one and copy it to sell at 25%. He can make a profit at a lower threshold than you, so you as a creator cannot compete.

So don't compete by selling copies but by funding the creation up front. No one is claiming that abolishing copyright won't be disruptive to any existing business models - in fact, that's the point: once something becomes part of our shared culture it is ridiculous to let one entity continue to have exclusive rights so if your business model relies on continued royalties, find a better one.

Otherwise, perhaps consider continued payments to everyone who built your house, computer and whatever else you use if you think that is a great way for society to function. Don't worry, the way things are going we might get there via technical means anyway.

But what do you mean, no incentive to create?

The Tao Te Ching was reluctantly written after the author was begged by his pupils. Most Greek philosopher's teachings were only written down after their death because other people thought that's an important job. On The Origin Of Species is a book because that was just the normal way to communicate scientific findings in Darwin's time. Da Vinci saw some fat commissions in his life, but Mona Lisa certainly never brought him any money. In fact, out of my twenty favorite artists maybe two saw anything approaching fame in their lifetime.

Please, go to some random DeviantArt page or Spotify profile or GitHub repo with 3 views and tell me why it exists when the only reason for human creation is dollars and red carpets...what a sad perspective, really

That doesn't mean I must be in favor of every repressive innovation-stifling law that was ever cooked up.

You bring up the arts in another comment; ever considered why like half the people regarded as genuinely world-changing or geniuses (da Vinci, Galileo, Columbus, Machiavelli, Michelangelo) were born in the same two hundred years in the same region? Because the Italian renaissance was all about intense, free information-sharing! People freely visited each others work places and ruthlessly stole form each other, and it was accepted. Boom, you get a period of unparalleled human productivity.

And now you want to tell me that a set of weird laws who only ever benefited Disney and Elsevier are the only thing preventing humanity from ceasing to create awesome shit? Nah man, the masses will always continue creating, exactly as proven by the fact that they did in the last decades while getting continuously butt-fucked by the very laws you pretend are made to protect them...

2) So far, these tools are "better search" schemes, not actual intelligence. Sure, many find them very useful. But given this, the (voluntary or involuntary) providers of data ought to get credit/benefit for/from this phenomena, along with the tool creators. Especially giving the current situation is Microsoft/OpenAI selling to commercial software developers who sell to general public.

What they are good at is predicting what's after the text. The problem of predicting what's next could be used to create a universal artificial intelligence (there's a mathematical definition for this). I.e. if you have a system which is very good at predicting what's next, you could get to very powerful AI.

If you are interested, you could read about it here: http://www.hutter1.net/ai/uaibook.htm

The intelligence of human beings isn't unspecifically good "predicting what's next" but rather is good at particular sorts of predictions in particular contexts, often involving the person having helped create the situation. I'm fairly safe at driving because I maintain an arrange of my vehicle in a fashion that allows me to predict easily what's next as well as allowing me to adjust if my predictions are wrong. Self-driving software might predict what's next as well as me in normal circumstances but it's neither aware of larger context nor does it things to maintain "smooth traffic flow".

Opposite, being able to predict anything generically would certainly be limitless intelligence but you can't describe any system with just that. Copilot is trained with a certain window, with the transforms special element giving more context but I don't think very many people doing current research expects that become generic prediction. I think I'm describing the consensus that it's a "better Google" for finding code one can use - and Google is a pretty good resource for this - if you aren't doing something unusual or difficult.

Jeff Hawking also makes "prediction is intelligence" claim but I think your and his approach misses that human intelligence is good not by being generic but doing more specific things.

I don't see a world where Copilot isn't fair use (at least in America).

-Some universe out there, with a god smarter and better than ours-

The situation with Microsoft and Copilot is the exact opposite. Here, Microsoft is misusing its acquisition of GitHub to repackage the work of individual free and open source contributors into a proprietary product in violation of the authors' software licenses. These licenses do not even require Microsoft to pay. They only require attribution and redistribution under a compatible license. Supporting Microsoft's misuse of GitHub is an anti-populist stance that puts the interests of the corporation over the interests of the individuals.

In the case of copilot, the damage suffered by the authors is close to zero. And those who benefit the most are the authors themselves. A double digit percent productivity enhancement is worth more to me than a few million $ to a trillion dollar company, especially because MS has to pay for compute.

Microsoft can even continue to sell Copilot as a service while keeping it license-compliant, since most developers are not going to self-host the entire dataset. Microsoft can also choose to exclude copyleft-licensed code from Copilot or create multiple flavors of Copilot, each licensed differently. You can get your "productivity enhancement" without needing Microsoft to violate software licenses.

The damage is not in the monetary payment denied to free and open source software contributors, payment these contributors never demanded. The damage is in Microsoft violating other people's software licenses to create a proprietary product derived from copyleft-licensed and attribution-required code, and in Microsoft encouraging other developers to violate these licenses. Microsoft needs to rectify these violations with specific performance.

Why can't they train on the code they own, such as Windows sources for example?

Or even better, why can't they release CoPilot itself under an open source license that is compatible with the licenses of code they would like to train on?

Also, I don't think anyone cares about the monetary aspects. The idea behind the GPL style license is to make sure that code remains free, regardless of what or who uses it. Freedom in this context refers to the ability study the code, modify the code, and distribute any modifications. Without the GPL the code can be used in a proprietary product which strips those rights away from users of the product.

Copeleft uses copyright laws to attempt to guarantee freedom for users. This is the inverse of what normal copyright does, which is allow a single entity to sit on the ideas and not allow other's to benefit from them.

If we can just strip copyleft licenses from projects, we are giving up those guarantees that GPL code will remain free for all users.

The GPL is trying to do it's job here, not slow down progress. Progress would be everyone benefiting from the technology behind CoPilot, rather than just MicroSoft sitting on the project and selling it as a service.

Copyright working in a supported/non hated way: You develop a package to do X by cribbing off someone else's package X. They sue you for stealing their work, not to make money off you. Situation at hand is case 2, hence the lack of interest in financial gain.

Why is this case 2, when it does not always reproduce the copyrighted works exactly? Situation: You realise that rather than cribbing off of one persons package X, you can crib off two other package X's and mix/average their contents. Scale this to 100's of packages.

Eventually, ML should avoid this by developing to work from first principles, writing in it's own style, with public code used only for validation of it's ability to understand and write code.

I actually agree. However this is not what's happening here. Copilot effectvely removes copyright from FLOSS code but doesn't touch proprietary software. FLOSS loses it's teeth against the corporations.

This is ridiculous; we created A is that can program themselves and people are worried about incidental copyright infringement.