GioCities

blogs by Gio

Tagged: enforcement

šŸ–± So you want to write an AI art license

  • Posted in cyber

Hi, The EFF, Creative Commons, Wikimedia, World Leaders, and whoever else,

Do you want to write a license for machine vision models and AI-generated images, but you’re tired of listening to lawyers, legal scholars, intellectual property experts, media rightsholders, or even just people who use any of the tools in question even occasionally?

You need a real expert: me, a guy whose entire set of relevant qualifications is that he owns a domain name. Don’t worry, here’s how you do it:

This is an extremely condensed set of notes, designed as a high-level overview for thinking about the problem

Given our current system of how AI models are trained and how people can use them to generate new art, which is this:

sequenceDiagram
    Alice->>Model: Hello. Here are N images and<br>text descriptions of what they contain.
    Model->>Model: Training (looks at images, "makes notes", discards originals)
    Model->>Alice: OK. I can try to make similar images from my notes,<br>if you tell me what you want.
    Curio->>Model: Hello. I would like a depiction of this new <br>thing you've never seen before.
    Model->>Curio: OK. Here are some possibilites.

The worksšŸ”—

The model and the works produced with the model are both distinct products. The model is more like processing software or tooling, while the artistic works created with the model are distinctly artistic/creative output.

Models do not keep the original images they were trained on in any capacity. The only keep mathematical notes about their properties. You (almost always) cannot retrieve the original image data used from the model after training.

sequenceDiagram
    Curio->>Model: Send me a copy of one of the images you were trained on
    Model->>Curio: Sorry, I do not remember any of them exactly,<br>only general ideas on how to make art.

There is a lot of misinformation about this, but it is simply, literally the case that a model does not include the training material, and cannot reproduce its training material. While not trivial (you can’t have a model if you can’t train it at all), when done properly, the specific training data is effectively incidental.

AI-generated art should be considered new craftsmanship — specifically, under copyright law, it is new creative output with its own protections — and not just a trivial product of its inputs.

PlagiarismšŸ”—

The fact that AI art is new creative output doesn’t mean AI art can’t be plagiarism.

Just like with traditional art, it’s completely possible for specific products to be produced to be copies, but that doesn’t make that the case for all works in the medium. You can trace someone else’s artwork, but that doesn’t make all sketches automatically meritless works.

The inner workings of tools used in the creation of an artistic work are not what determines if a given product is plagiarism, or if it infringes on a copyright. Understanding the workings of the tool can be used in determining if a work is an infringement, but it is not the deciding factor.

šŸ–± Ethical Source is a Crock of Hot Garbage

  • Posted in cyber

There’s this popular description of someone ā€œhaving brain wormsā€. It invokes the idea of having your mind so thoroughly infested with an idea to the point of disease. As with the host of an infestation, such a mind is poor-to-worthless at any activity other than sustaining and spreading the parasite.

A ā€œpersistent delusion or obsessionā€. You know, like when you think in terms of legality so much you can’t even make ethical evaluations anymore, or when you like cops so much you stop being able to think about statistics, or the silicon valley startup people who try to solve social problems with bad technology, or the bitcoin people who responded to the crisis in Afghanistan by saying they should just adopt bitcoin. ā€œBad, dumb thingsā€. You get the idea.

And, well.

Okay, so let’s back way up here, because this is just the tip of the iceberg of a story that needs years of context. I’ll start with the most recent event here, the Mastodon tweet.

The Mastodon ContextšŸ”—

The ā€œheā€ Mastodon is referring to is ex-president-turned-insurrectionist Donald Trump, who, because his fellow-insurrectionist friends and fans are subject to basic moderation policies on most of the internet, decided to start his own social network, ā€œTruth Socialā€. In contrast to platforms moderated by the ā€œtyranny of big techā€, Truth Social would have principles of Free Speech, like ā€œdon’t read the siteā€, ā€œdon’t link to the siteā€, ā€œdon’t criticise the siteā€, ā€œdon’t use all-capsā€, and ā€œdon’t disparage the site or usā€. There are a lot of problems here already, but because everything Trump does is terrible and nobody who likes him can create anything worthwhile, instead of actually making a social networking platform, they just stole Mastodon wholesale.

Mastodon is an open-source alternative social networking platform. It’s licensed under an open license (the AGPLv3), so you are allowed to clone it and even rebrand it for your own purposes as was done here. What you absolutely are not allowed to do is claim the codebase is your own proprietary work, deliberately obscure the changes you made to the codebase, or make any part of the AGPL-licensed codebase (including your modifications) unavailable to the public. All of which Truth Social does.

So that’s the scandal. And so here’s Mastodon poking some fun at that.