The AI elephant in the room

Lots of people are upset and disappointed at me for using AI, leading to much speculation online.

History
My first use of AI was in 2020, a piece called “On Hold for You”.
Much of that music was algorithmic, and AI helped make the first realistic saxophone I’ve ever produced.

Since then, I’ve tried nearly every new AI I could find.
I had spreadsheets rating dozens of them, what they did well, and how they did it.

I’ve offered (and still do offer) my music to anyone who is training up a music AI. Several researchers took the offer, and I couldn’t be happier.

Using AI
Today, the music AI I use most is Suno.
Suno generates endless blues songs in whatever style I want. I use these to practice harmonica.
I use Suno to make relaxation and motivation music that I can’t create any other way. I listen to these at my house for relaxation and motivational purposes.
It helps create quick backing tracks for lame JavaScript games that I enjoy making.

I don’t publish any of these on incompetech. They don’t get distributed to music stores or streaming services. I have posted a few cool ones to freepd.com, my public domain music site. I’ve been very up-front with what these are, and how they were made.

So yeah, I use AI. I find it incredibly powerful and fun.

Motivation
The reason I wrote the music I wrote is because people needed it. I make utility music, without words, not to express myself, but to express feelings that people needed for their games, videos, and movies.

In recent years, people have needed new music less and less, so I haven’t really been motivated to publish more.

AI doesn’t work for everything yet. So, I still have a few human-made pieces in the pipeline.