I think their business model is to work on specialized models rather than general purpose. See https://pleias.ai/blog/sillon-ratp and the focus on synthetic datasets for specific personas.
I don't know if it's ethically better to use LLMs trained on data licensed from X, Reddit, stackoverflow, Sony, CNN and all big content aggregators who will/have agreements with big tech. I'd prefer to focus on mechanisms to force reciprocating the donation: scrape and train at will, publish the models as open weights, at least.
Anyway, the vegan LLMs exist, see the work of Pleias.ai.
You don't need a license to scrape the public web and analyze it, turn it into tokens and other transformations. Let's not expand copyright beyond the horrible monster it already is.
You don't need to have fully copyright-unencumbered datasets to build Open Source AI, as that (as you say) would be impossible. https://opensource.org/ai
Unix workstations had mice with 3 buttons. The Mac had only one. Windows, Amiga, Atari had two. The Unix developers had choice that others didn't have. They came up with a use that existed forever. Now someone decided to remove the default for no apparent reason. It's like the Android product managers continuing to change the color, size and gesture to answer a phone call: every release, the first call is an exercise in managing frustration.
Mr. Martin was also paid to support the production of GoT, not just royalties. There is no reason to believe that he wouldn't be called to do the same sort of consulting work on the script, dialogues, visual, etc if the copyright expired.
Yeah... Ask Schrems about the hefty fines and all that pretty things bright to Europeans by the GDPR. Come on! The GDPR is at best a pretty face to a rotten nothing-burger.
Same as today? Empirically demonstrated: The only ones getting richer and richer after the Napster wars are the publishers, like Apple Music, Spotify and the other mega corporations. :)