I assume that the artists style is essentially just mapping from their name into some latency space of styles, more or less, not necessarily that the model has

I assume that the artists style is essentially just mapping from their name into some latency space of styles, more or less, not necessarily that the model has memorized every individual's style. If that's the case, it would be interesting to see what happens if the AI purposefully isn't trained on art in some part of the style latent space, and to see if it's able to still generate art in that style. Taking this to a possible extreme, what if we trained it only on classical art work, would it be possible to get it to extrapolate to more modern styles, or to fine tune it with only a small number of, potentially, as people are calling for, paid samples?