Machines cannot "reuse" things.

I strongly contest the fact that LLMs reuses things on their own.
The people making them give them the things they use, not the other way around.
The people asking them results expect them to reuse things, not the other way around.

Can you get me a single law decision stating that machines reuses things on their own?

tl;dr Your users want you to FIGHT for them, not lay down in the face of a crisis. This is, as the report notes, a "watershed moment" - what will you do? Sacrifice the users (and your brand...) for the right of major corporations to exploit and pollute the commons?¹

I have read the report. The fundamental issue here is this:

• CC has (correctly, I assume, IANAL) determined that the current CC license(s) are legally unable to protect the commons against exploitation by AI corporations
• The organization does not want to extend the licenses further, because they believe this to be an expansion of copyright, which must be defeated
• Therefore, Signals is an attempt to ask nicely that companies stop exploiting resources (artistic work, electricity, server bandwidth) while doing nothing to address the issue of author's rights.

Who does the organization serve here? Is it the adopters of the CC license? I would have assumed that to be the case beforehand, but now I'm not so sure. I can't help but compare with the FSF, which - when faced with existential threats like TiVoization and cloud services - responded to protect its users by issuing new versions of the license (GPLv3, AGPL) that closed such loopholes. There was no attempt to put a special HTTP header on Github signalling "hello Broadcom, please chip in a couple bucks back to the Linux project". There is in fact an entire legal operation in the EFF devoted to lobbying and lawsuits to protect those who adopt such licenses. And yet, it thrives - the ecosystem is healthy (ish), and we can still get nice things like "allowing for translation" in FOSS software, despite such "hard" tactics!

Has CC even attempted legal recourse against purported violations? The report says, for example, "Litigation involving enforcement of CC licenses is rare, and much of it involves litigants who are not operating in good faith." Do you really believe that? If I sue someone for sanding my name off an image and putting it in their multi-million-dollar video game, am I Bad Faith in the eyes of the organization, because I dared to invoke my rights? It's not surprising, I guess, that the organization would throw up their hands and suggest how best to roll over rather than get a lawyer, if that's thought to be simply beneath them.

Well noted regarding machines not 'reusing' things on their own. Will aim to veer away from that language going forward.

I've also read the comment about users wanting CC to fight for them. This is definitely a theme we've seen from the feedback thus far and thanks for laying it out so clearly here.

Thanks for this invitation! Our vision for CC signals is grounded in CC’s mission to empower people to share knowledge and culture openly and for the public good. Our goal is that this tool helps consumers know which AI developers are more ethical actors by adhering to CC signals, which would in turn drive internet users to choose more ethically created tools. Through building better norms around AI development where AI developers are always giving back, always providing transparency and attribution in whatever ways are technically practical, always making their models available for others to build upon and use. When signals become part of everyday practice, creators, communities, and institutions feel that they can, in good conscience, contribute to the commons. And AI developers feel real pressure to meet higher standards.

The outcome would be that 1) the benefits of the commons are more evenly distributed for the public benefit, not just for profit; 2) when using AI tools, consumers are able to choose those that better align with their values and interests and 3) lawmakers see that clear, community-driven standards are possible, and set higher expectations globally for AI developers that support the public interest and the commons. Simple standards for giving back can be set and adhered to.

This vision is about more than one tool. It’s about shifting norms and incentives so AI gives back to the open ecosystem instead of just taking from it. This also aligns with a recent report from Open Futures where they call for a fairer redistribution of value to keep the commons thriving.

Ten years from now, our goal is that CC signals will have helped shape a world where the benefits of the commons truly flow back to the people and communities who keep it alive. Where creators feel confident to share, knowing their choices matter. Where people can choose AI tools that respect human intent and contribution. And where technological progress serves the public interest (not just profit).

@ccjocelyn thanks for the reply! I applaud the aim to foster sharing knowledge openly and for the public good and I believe your intentions are that the Signals proposal shall help towards this aim.

However, my question was aimed more towards a clearly defined outcome in ten years time, and how a stringent explanation could look like that logically and coherently leads from the status quo, through the Signals, to that desired outcome.

Right now, the explanation you provided looks a bit like this to me:

1. Signals are introduced
2. (something magical happens)
3. a brighter future for the Commons

This looks at odds for me in several important categories. I have already outlined them in a lenghty blog post that was linked in the discussions here. I will try to boil them down to a few questions:

How can you ensure the mechanism won't be gamed, leading to a failure of your outcome 3 (lawmakers see benefits)

The proposal, as it is, seems to incentivize actors to flood the web with supposedly their own creations (or machine creations) that they can then outfit with Signals in order to redirect remuneration to a sink of their choice. How would this be addressed? Since remuneration plays a vital role in two of your four Signals, this seems to be a core aspect. This looks like the number one issue that just begs to be Monkey Pawed.

How do you deal with the cultural shift of what Commons are?

I have always understood Digital Commons as something where everybody contributes what they can, to the benefit of everybody – since they are non-rivalrous, it doesn't matter whether 99,9% of all benefitors never contribute to the commons. It could also be said that a way towards more Commons for everybody would be more free time for everybody. Or having their basic needs met. Projects based on Commoning practices were, for me, always examples of how we could go about producing stuff for the benefit of all, without remuneration logics attached to them.

With your proposal, you seem to turn this picture upside-down. You stress in your explainatory PDF that the remuneration is not meant to be a commercial transaction. Both remuneration Signals are, however, value transactions. Suddenly, the idea of the Commons is not tied to providing value that is not meant to be appraised in financial terms anymore, but it's about explicitly putting a dollar number to it. You might say that you're just one actor, but alas, you carry the term “Commons” in your name.

What I take from your described goals is: In your proposed future, “distributing the benefits of the Commons” is tied to actual financial transactions. Starting from the problem at hand, you might also lobby for taxing genAI vendors in order to distribute the current wealth a bit better – which might enable more people to contribute to the Commons. But you decided that financial transactions should be the way. How do you deal with working towards a future in which the idea of Digital Commons are so closely tied to financial transactions, turning them into commodities?

You also use the term “enclosure” in a very interesting way. Instead of framing the creation of generative models as enclosure, in which a few actors abuse the Web for creating products only they can produce and keeping them closed, you put the burden on actors who chose not to contribute to the Commons anymore. Also including journalistic outlets (who rarely really published Commons) who are putting up paywalls – as if this were a direct result of current genAI practices instead of a process that has been ongoing for about a decade now.

Generally speaking, I completely miss any reflection on power structures and hierarchies. Your proposal looks as if all involved parties were operating on equal footing. Why is that?

What best way of actually putting this into practice do you have in mind?

Let's assume we have all these remuneration and crediting aspects. The big question for me is: How? What are your specific, tangible ideas for a way forward in order to get to that outcome?

Will we have devices in which we will not only be able to read through 1500 pages of F/LOSS licenses and credits in case we are bored on a slow sunday afternoon, but also through 3000 pages of Signals credits? Whom would that actually benefit? Would others have to make sure that vendors don't just generate a plausible list of credits in order to look “ethical” to consumers as you outlined in your outcome 2)? Would that not just lead to drudge work for maintaining the list and others to validate it, if that is even feasible due to link rot and all the other practicalities of the Web as it exists in reality? Just more mindless bs jobs instead of, you know, spending that time for contributing towards the Commons?

Also: How would the collection of the remuneration work, in practice? Do you intend to resurrect the Micropayment ideas from some 15 years ago? Or collecting agencies tied to the Commons, also an idea that had already been tried in the early 2010s? What would be your preferred way to implement this?

And what structures and dynamics would that lead to? And which of those would you say are welcome? Publishing creative works on the Web that can even be exploited in the training of genAI is, globally speaking, a privilege of a wealthy few. Will those few have the added privilege to be able to funnel some of the funds of some VC-backed entities into a cause of their choice? What is your idea how this is supposed to tie into all the other sectors that are affected by the current genAI hype? Do you have ideas how this should end up coming together in ten years time, or is this a case of “not my department” and whatever outcome is okay?

Hi @stkdiretto! Really appreciate all of this thinking and curiosity. While I'm thinking about a few of these pieces, I wonder if maybe you'd want to chat about it? Feel free to reach out - jocelyn@creativecommons.org.

Really appreciating this whole thread, @stkdiretto. You are poking holes in exactly the right places. No easy answers for you right now but your point about needing to clearly define the outcome is well taken. We're keen to keep chatting about this in more detail.