AI can generate sound, but only a person can decide what the work should express — let human response give it meaning; discovery starts there.