When a pile of coalitions go "this new bill is awesome" and not a single one is any science or tech group, be a tad suspicious. But I was telling AI folks "ingredient lists" were coming in 2022.
@textfiles Wild! Effectively requires the creation of a comprehensive #copyright register, even for works the authors themselves have not bothered to register.
@trashheap Using a built-in software feature, like “reshare”/“boost”, is not a #Copyright infringement. I can't remember which case it was, but a high court in the USA explained the different between re-uploading/redistribution vs. using a platform's built-in “reshare” feature.
What can be considered as infringement are:
Re-uploading / redistributing an image/photo/art/fiction/ebook/story. Since these are automatically Copyrighted to the author/creator/writer.
Creating a derivative, or a reproduction, work without permission or a license.
On screenshots, it is case-by-case:
Was it a screenshot of a scene from an animation or live-action production? It is infringement, however, in countries where there is Fair Use, or similar, they can get away with it depending on usage. (Example, in an article or review, it falls under Fair Use. [But if you overdo it, you still might get in trouble.])
Was it a screenshot of a post where a Copyrighted work was included? If the subject is clearly about the post and not the Copyrighted material, it is fine. Otherwise, it's a grey area. If the owner of the Copyrighted material that was included in that screenshot, can prove their Rights were infringed, they can potentially win in court. (But using a platform's “reshare” feature is, again, fine as you simply re-shared it, not re-uploaded.)
How about text-only posts?
Generally, text-only posts in microblogging platforms, are not Copyrighted as those do not fall under any conditions of the Copyright Law. If the post owner can prove their post was a creative work, then they can try it in court.
Of course, the Copyright landscape is constantly changing in our fast evolving cyberspace. For example, in many platforms today, they allow threading of a post. In some, even though the platform doesn't have any threading feature, mobile apps can add such features (example, Trunks can display a Mastodon thread properly even though Mastodon's web and mobile interface don't support threading).
So, if a user can prove that it is a “thread”, then there is a chance it can be Copyrightable. But, again, largely, “no”.
Anyway, it's the simplest way I can explain it. Also, #IANAL and #TINLA.
Every time, you think it couldn't get any worse, a new revelation tops it off. As an author, I wonder how long it will take for the book market to be completely enshittified.
Thank you for the #giftArticle! ⬆️ @writers@bookstodon
TL;DR: #Copyright is not a fix for your #AI fears. Bullies want you to think they're on your side, but it's a trap. Bosses blame immigrants for low wages, while they actually take workers' wages. Corporations claim to protect creative workers from AI companies, but they exploit them just as much. It's a ploy to gain control and exploit workers, while deflecting blame onto others.
Always keep in mind there is
a legal difference between the content you post publicly and the content you post using a service implying you have agreed to its Terms of Service.
Always remain aware of what you publish where, and what the Terms of Service and Privacy Policy of this service are.
lmao. I got a DMCA takedown notice for my Colordle game yesterday ( https://colordle.lina.garden ), along with everyone else who forked the Reactle repo. today 404 is reporting on how bullshit it is! 🤣🖕
More Than Just a Game #law conference at Stanford Panel 4: Fan Protection. Panelists are Ben Golant - Tencent America, Joseph Schenck - Fenwick & West, Nari Ely - Epic Games, Jordan Gimbel - Twitch, Michaela MackDonald - Queen Mary U of London, Chris Cotropia - U of Richmond Law. #gaming
Dunning: When I litigated the Monkey Selfie case, an important element litigated was something that was missing from the #copyright act: human authorship. In that case, the courts decided that human authorship was required, but the court didn't define what was required to establish human involvement in terms of all the different roles in #ai
If copyright is real then the AI machines owe the artists of humankind (which is to say all of us) more money than exists in the banking systems of the planet.
If copyright is not real then the companies that trained those models have no right to exclusive ownership of the weights.
I bet somehow the law conspires to make it the worst of either option.
According to DocuSign’s website, DocuSign - a company most commonly used for helping people sign contracts digitally - is or may be (its information is not wholly clear) using your contracts and other documents to train AI.
I didn’t know this, but then I don’t use DocuSign, although some of my clients do.