02 · AI Music
AI Music
I got into generative music early — the first weeks Suno was public — and have kept at it since, to the point of several full albums distributed through DistroKid. This page is about what I’ve made with it and how I think about it.
“Blind Visionary”
One of the first tracks I made when Suno appeared. A rap song, which was not the genre anyone who knows my listening habits would have predicted. That was rather the point: the interesting thing about a generative tool is that it lets you work in a style you cannot personally play.
Generated with Suno AI, early in its public release. Watch on YouTube (opens in a new tab).
The albums
Several albums of this material have gone out through DistroKid (opens in a new tab), which means they are on the usual services — Spotify, Apple Music, YouTube Music, and the rest of the list a distributor pushes to. The YouTube Music side lands on an auto-generated artist topic channel, which is the tidiest single place to browse the releases in one go.
- Albums on YouTube Music (topic channel) (opens in a new tab)
- Hurst Sound Studios (main channel) (opens in a new tab)
Searching Hurst Sound Studios on your streaming service of choice should turn up the same catalogue. If a service has it filed under something else, tell me and I’ll link it directly.
Where I land on all this
I don’t think a prompt makes anyone a musician, and I’m not going to pretend the tool did nothing. Both of those positions are easy and both are wrong.
What generative audio has actually been for me is a way to hear an arrangement idea in ten minutes instead of ten hours, and to work in genres I have no business trying to play. That is genuinely useful. It is also, plainly, a different activity from sitting behind a kit — and this site keeps the two on separate pages for that reason. Everything on this page was generated. Everything on the Recordings page was played, programmed, or mixed by hand.
What I’m watching
The part of this field I actually want to understand is underneath the prompt box: how these models represent timbre, and what the same techniques do for instrument modeling and sound design — where the goal isn’t a finished song but a convincing instrument you then go and play.