I read someone saying that A.I. music has arrived, and it is in the form of Beach Boys Sing the Beatles. They like listening to it, so it is real art according to them.
I gave it a shot. To me, it is a computer novelty. If it was on at Trader Joe’s, I’d say what is the glurge and try my best to focus on the groceries.
It is very good as a simulation, but it is lifeless — more lifeless than the KPM library music that was intended to be lifeless. Far more lifeless than entirely computer-generated music made from human-designed algorithms like Laurie Spiegel’s.
Though there is no denying that this may be what people want from music. Accurate simulation of things they already like.
Which puts me back in a state of mind I was in high school: completely not understanding what people wanted out of music and why popular music was popular. Except this is on another level. Warrant and Pearl Jam and Boyz II Men had an appeal that escaped me, but it wasn’t dead-sounding like this A.I. mashup.
When I was working at Spotify, I was thinking about the data indicated that the best way to maximize streaming was to never draw attention to the stream by throwing in a track that was jarring and/or unfamiliar, and I thought, what if the best way to do this was not with songs at all, but rather familiar songoids blended in a smooth stream. Eventually, that stream could become self-similar, rather than similar to music, becoming comfortable audio stimulus instead of music.
I also futilely argued a few times that total minutes spent streaming wasn’t the way to measure satisfaction with the service. What if a user heard something really striking for ten minutes, then stopped listening for the rest of the day because they wanted to digest it? Wouldn’t they be likely to continue their subscription?
I may have been wrong, and of course, I was laughed off when I suggested this. But it looks like commercial music is going to shift even further from that idea of musical value and more toward comforting sounds.