Post by @ljwrites@writeout.ink

To distinguish: Copyright is the exclusive intellectual property right, usually limited in time, to copy, distribute, adapt, display, etc. a creative work as a specific expression of ideas.

Moral rights are the related but separate rights of copyright holders to recognition for, anonymity in, and integrity of their works. They are separate from intellectual property rights, and attach to the person's inherent dignity and integrity — so much so that in jurisdictions such as Japan, China, and Korea the term was translated as 著作人格權, "personhood/personal dignity rights in creativity."

0 Boosts

0 Likes

Author-ized L.J.

@ljwrites@writeout.ink 8 hr. ago

When scraping, non-attribution, and plagiarism by machine-learning (ML) models are at issue, the question is not only one of copyright and other intellectual property but of the personal/moral rights of creators as well. If a scraper takes a mathematician's work, and a model trained on this data uses it in a proof without citing the mathematician, the person's moral right to attribution is implicated in addition to copyright. If a human did this we would call it plagiarism or at the very least a poor citation practice, and I don't think anyone other than the most disingenuous and ruthless of AI boosters is willing to die on the hill that it's fine for ML models to output results that would be improper if a human had created it.

Author-ized L.J.

@ljwrites@writeout.ink 8 hr. ago

Also, while in a strict legal sense moral rights may not be enforceable in every jurisdiction in the same way or to the same extent, and the usual statutory implementation is for copyright holders, I think we recognize the ethos beyond jurisdictional differences and the strict wording of legal texts: That giving credit to creators and preserving the integrity of their creative works is not just about obeying the law but about cultivating creative, scholarly, and professional communities by recognizing creative contributions, and through it the creators' membership and standing in these communities. By using people's works without consent and cutting the creators out of credit, machine learning systems deny creators their agency and contributions, with implications for their dignity as creators and as people. As the terminology for these rights imply, these are MORAL rights about community ethics and the dignity of creators as people. These are not issues that can be swept under the same rug as objections to copyright as property rights.

Author-ized L.J.

@ljwrites@writeout.ink 8 hr. ago

As a practical matter and also separately from copyright itself, receiving citation and credit directly implicates many creators' reputations and ability to pursue their work. Scholarly fields have formal citation metrics that form the basis of evaluations, grants, and reputation. Even when the metrics are not so direct and measurable, being recognized as the founder or major figure of a field, style, or genre can give creators a whole new level of standing and influence, like Alphonse Mucha with Art Nouveau or J.R.R. Tolkien's influence on fantasy. Moral rights and adjacent ideas/practices have huge impacts on the lives and even afterlives of creators, from livelihood to legacy. These are community-building ethos that may be damaged by scraping and non-attribution, and the call to "abolish copyright" cannot adequately answer any of these concerns.

Author-ized L.J.

@ljwrites@writeout.ink 7 hr. ago

I will conclude for now by briefly touching, as promised, on the copyright issue I set aside at the start. As discussed above, many issues with scraping and attribution practices do not even concern copyright. But even if we supposed there are cases where where nothing but copyright is implicated, I question the intellectual honesty and proportionality of bringing up copyright as a defense to scraping, particularly if the only or rare instances where someone brings up the evils of copyright are where copyright might be a legal obstacle to scraping.

If the most strenuous objections to copyright are not brought up in the hoarding of wealth by billionaires and detriment to the culture, but rather where it is suggested as a minimal and imperfect protection for individual creators, scholars, and professionals to receive credit and recompense, I will be inclined to question whether what is being expressed is a principled objection to copyright or a gotcha to defend ML scraping and plagiarism practices.

1 more reply