In a way, the “it just does what humans do but faster” argument is starting to follow the “a number can’t be illegal” trajectory.
Sounds like the AI model should be paying royalties to every affected artist for the right to sample their work.
> At CC, we believe that, as a matter of copyright law, the use of works to train AI should be considered non-infringing by default, assuming that access to the copyright works was lawful at the point of input.
https://creativecommons.org/2021/03/04/should-cc-licensed-co...
It's not backwards. It's the same as if a human artist studied it.
Backwards would be thinking that any creative work you make is a derivative work of many creative works that you've seen in the past which you aren't copying from.
It’s not a human artist, and it can only regurgitate mash-ups of work stolen from others.