> To the best of my understanding of the law, making whole copies of a work for other people to then make fair use of is generally not considered fair use.
Who is making the copy? You have your copy that you bought. Bits in a wire are transient so that shouldn't be a copy any more than opening a book and exposing it to a reading lamp. The copy is made at the destination by them on their own storage medium. It still seems hard to distinguish from borrowing a book and making fair use of that.
> Likewise the person downloading it can't be making fair use of the work unless they already have a copy of the work they have legally obtained. This was a key point of the Anthropic case, the use of legally purchased or acquired books was fair use, the use of the copies obtained only via piracy were not.
Which is the thing I'm struggling to understand the logic of. It's like calling it piracy by stipulation from the outset and then using that to decide if it's piracy or not. That's just assuming the conclusion. If the person uploading it has a purchased copy, and the person downloading it in order to make a copy is making fair use, where is the piracy happening? A fair use copy of a lawfully purchased work is being made. It seems like they're just trying to find a way to call it piracy because it involves file sharing without even considering what's actually happening.
> Libraries have been something publishers have hated for a long long while.
Well yeah, but they're not supposed to be the ones who decide what the law is.
> it's unlikely you can convince a library to let you just scan all of their books regardless of the legality of you doing so.
Isn't scanning books from libraries the thing Google Books did? And the legality is the issue.
> As for the partnership thing, my point was mostly that any legal entity that can jointly own property would have to be the owner of a work for people within that entity to distribute copies between each other, and that once a person leaves the legal entity they could no longer keep or obtain copies of works owned by the entity.
Which still doesn't seem like it quite works.
Suppose you want to train an AI model, so you buy a bunch of books, scan them, train the model and publish the model on the internet. Presumably someone who gets a copy of the model doesn't need a copy of the books? But then why would you need a copy of the books? Couldn't you discard the scans, sell the books and keep the model?
At which point, how are we distinguishing this from borrowing the books, training the model and discarding the scans?
We can also consider this from a policy perspective. If Google had to buy a single copy of the book and wasn't allowed to sell it, would that really matter to the publisher or Google? Nope. The publishers are just looking for a hook to sue them. But it would matter to someone smaller who was trying to do it on a budget because it would make the smaller or non-profit effort non-viable, which doesn't get the publishers anything either way. So what should we do here? Probably not the thing that gives a big advantage to massive incumbents over their smaller/decentralized/non-profit competitors without providing the publishers with anything of significance.