I appreciate the detail in your reply. Do you think the recent Warhol "Orange Prince" case gives an inkling into possible future court treatment of the question of "transformative" use for generative AI models? There Warhol's silk screen print of the original Prince photo was deemed not transformative enough as I understand it. One of things about the stochastic nature of generative AI is can be rather hard to notice when the model spits out something very close to the training material.