Hmm, I don't think the art/media examples really work as well here compared to the scientific discovery ones. Whether a piece of art or fiction does well doesn't really correlate that heavily to any objective form of quality in that work, let alone something that can be entirely replaced by automation.
I don't see how AI means that something like Hollow Knight: Silksong or Donkey Kong Bananza wouldn't do well. What is the actual reasoning behind say, the next indie game or movie not being appreciated because AI can design or create games or movies?