If I can put in a mood or choose a set of requirements (like in Garageband) and out comes what I want to hear, that’d be fine with me. Some Sunday mornings, I just want generic “cafe music”, a genre that could likely be VERY faithfully copied.
In reading
“The technology is able to generate dynamic soundtracks that change based on user interaction.”
what I actually get from that is something that Apple could tie to Scene Kit. Where any application that wants to take advantage of generative audio without creating their own engine could do so via a few straightforward commands that would describe the range and provides the parameters that could be used to control the “feel” of the audio.