I have no doubt that AI can make things- wonderful, perplexing and even beautiful things.
And the dispersion of individual authorship or artistic intent into the realms of chance, material processes or system-based outcomes has been well under way since Dada was born in the mid-twentieth century.
I believe your post about AI based art is hinged on fundamental misconception: art is not an object; art is a behavior. This is what I mean by cultural context; it’s done by humans, for humans, among humans. Art serves an evolutionary purpose; it’s a defining, behavioral aspect of our species, an integral part of our animal-ness. Could AI one day join us in this behavior, augmenting or spurring cultural development? Sure, but your examples are not this; they are pastiches of previous painting styles externally imposed on images. Moreover, style emerges from the interplay of materials, individual disposition and historical context, and I would argue that the unrepeatable, unique qualities of this interplay are what keeps us interested in objects well-past their specific cultural resonance. Again, maybe AI will one day generate a ‘style,’ but this AI will surely have to give us reason to care about it.
If you’re wondering why I would pick this bone, or why I privilege the evolutionary intelligence embodied by art as a species-wide behavior over the current state of AI, I direct you to Moravec’s Paradox. It’s implications for the development of AI are pretty straight forward, but I believe it also contains important conclusions about our conceptions of intelligence. The consciousness and abstract reasoning we currently possess are extremely young and late-to-the-game as far as intelligence is concerned (though very powerful indeed). So the ways we conceive of and design AI - and our most basic assumptions about ourselves- are inherently limited by this fact.
In other words: hubris.