Why now?

From 2000 until 2024, I wrote reams of music. Most of it was performed — by my high school band Good Stuff, by my college band Two Mikes from Jersey and the Exit 9 Band, by my a cappella group Fleet Street, by the Ram's Head Theatrical Society, by the Ensemble 101, by musical ensembles all over the world, and most recently, in video games.

Some of it, however, was never heard by a soul. When I lived in Belfast from 2005–2006, I already realized that this would be the case. I was 23 and writing at least a pop song every few months, and I knew I couldn't sing them myself. I also knew that the tug of my education in contemporary classical music wouldn't bring me to meet the right performers to bring this music to life. I resolved to simply keep making notes to self, keep humming, and keep writing.

As the songs came, I'd be in disparate places. Strung out on the couch of a friend in New York City. At a cafe in Nice. In front of a piano writing for a show. Editing an audio sample. I eventually organized them into five albums named after the streets I was on when I wrote them: Route 18, Lisburn Road, Palm Drive, Impasse de la Blonde, Sepetlahdentie. But with all of them, even if they were performed in some limited context, I knew as I was composing them that my ambition for them was much larger — I imagined orchestrations within contexts that I did not have access to and did not know how to chase.

In 2021, I started experimenting heavily with audio in AI. At Meeshkan, we built a state-of-the-art AI vocal artifact remover. I worked for a year at Audimee doing RVC voice transfer. I fine-tuned MusicGen and saw companies like Suno, using novel tokenizers and the transformer architecture, pull ahead in the data race, making more and more powerful models.

These innovations have had weird social consequences, many of which stem, in my opinion, from not understanding how to meet the current technological moment.

Record companies are in an epic battle with the frontier models and will perhaps settle for subsuming them into tools like Spotify. A massive amount of AI-generated music is uploaded to Deezer and Spotify — tens of thousands of tracks a day, which is obviously far more than humanity can consume, let alone appreciate. The tools are becoming more adept at covers and originals, and starting to resemble DAWs more and more, which extends the questions of ownership and rights directly to the most intimate choices during a creative process.

For me, this means that the recording is more or less dead. Not as an object of consumption — we will always want to listen to music. But as an object of value. Instead, recording will branch into two universes: (1) avant-garde artists, pushing the envelope of what's possible to express through sound, will find new palettes for us to explore; and (2) the recording will act as an invitation for collaboration. It allows us to expand our imagination of what a piece of music can be or become, which facilitates discussions between musicians and gives life to projects that would otherwise never see the light of day.

Raritan, this project, falls squarely in the second camp. I am not a talented enough performer to credibly interpret any of these songs on stage. But I love songwriting, and I feel lucky to have been born at a time when AI can translate detailed sketches into full-fledged arrangements that, while imperfect, represent musical thought in a way that words cannot. I'm using these artifacts, mostly produced by Suno, to connect with musicians and find happy homes for all of these songs. From singer/songwriters to big bands to glam rock groups, Raritan, for me, is about using a transformational technology — AI — to bring my dream of human connection to life.

This music has been a soundtrack to and for my life. With all of the trepidation and vulnerability that anyone feels when they share something intimate, it's my sincerest hope you like the music. If you're a performer and interested in any of it, I want to meet you, to hear your voice, to learn from you, and to bring these songs to life in a way that lifts you up the way this music has lifted me for decades.