AI can generate sound, but only a person can decide what the work should express — ask who gains access; attention must still be earned.