alt Hacker NewsThis post puts forward two paths:
1) Everyone and everything is subsumed into the forest. Innovation becomes unprofitable for the innovator as the one who controls the forest uses their capital to clone every new innovation.
2) Everyone withdraws from the forest. Innovation goes private. The forest stops growing, but doesn't die.
---
But there's two things the post doesn't consider:
1) Viral licensing.
What happens to a model if it is trained on data that comes with a license? What happens if the laws that be decide that the model producers, the models and the products of the models themselves must follow the conditions of the licences. How will that affect the model producers? What if customers don't want to be beholden to those licenses? What happens if the conventional wisdom is to avoid models to avoid lawsuits? What happens when models, model producers and customers power lawsuits against (other) model producers? Where would the new equilibrium between model producers and innovators move to?
2) Non-profits models
What happens to model producers if customer shift to become non-profits themselves, specifically those that pay employees instead of model producers. Would the model producers become starved out? Or would they need to switch to non-profit status as well? How would model producers, the models and the forest as a whole change if profit no longer became the priority?
Replies
> Viral licensing
IANAL, but for some months now I've been pondering what could happen if a site--like a personal blog--had a legally-strong "click-wrap" Terms of Service, which unlike GPL means it rests mostly on contract law, rather than copyright.
For example, imagine a ToS that says something like:
1. You acknowledge I am providing you something of value (my content) and you agree to provide compensation/consideration if you use that in an AI model.
2. By doing that, you grant me an irrevocable worldwide license to use and sub-license all content that emerges from the model, notwithstanding any future agreement you may make in reselling access or outputs of that LLM to anyone else. If a conflict should occur, you agree to indemnify me against claims by that other entity.
3. If you believe my content was not a material factor in some output, the burden of proof is on you to identify the specific output and prove that my content did not influence the it.
In short, this doesn't stop someone from stealing my art/writing, but it does put a potential hole in their attempts to monetize it.
For example, if they scrape my music and then license a copy of the new Music Generator 3000 to Disney, and Disney makes a movie with that music...
As the first line of the post says - it's a thought experiment, so comments like yours that open new options and ask new questions are the best outcome.
I have no other comment other than - very interesting. I thought about how the overlying model will change for us, but haven't considered that the underlying model (what you proposed) can change too ... if that makes sense.
what about we properly implement copyright and protection for software to prevent cloning style theft?
I mean we haven't had an innovation in patents and trademark for software for how long? Why is it that only hardware can be copyrighted and trademarked - can we really find no way to do this that can't be abused by patent trolls?
I asked recently on social media if anyone knows if there has been a legal decision regarding if GPL source code that was used for training an LLM will taint all that LLMs output with the same GPL licence. So far nothing has come up but I think people are wanting to know the answers.
It has been said that Microsoft indemnifies people using its LLM tools against copyright and patent issues, but I don't know if it applies to LLM output which might/should be GPL licenced.