“Had Meta bought plaintiffs’ works in a bookstore or borrowed them from a library and trained its Llama models on them without a license, it would have committed copyright infringement,” wrote plaintiffs’ counsel in the filing.

I wonder how many here agree with that theory.