A song is an abstraction of a time and place. Twelve years of releasing music taught me that the point was never the technique — it was capturing a particular version of myself before it moved on.
I release as Alter. under Alterxyourxego / AfterAlter.x, and produce for other artists.
How I actually work
My production style is brute-force iteration: constant layering and relaying, prototyping many versions of the same song like a multiverse of the same character arc, then choosing the take that tells the truth. That method is expensive in time. Audio AI makes it cheaper, which means I get to run more of it.
- Suno and Udio for fast melodic and arrangement sketches — disposable drafts that reveal what the song wants to be
- ElevenLabs for voice design and vocal texture exploration
- Whisper for transcription, lyric passes, and reference workflows
- Ableton Live and Logic Pro X where the actual record gets made: arrangement, mix, release prep
The line I hold
These tools generate options. They do not have taste, and taste is the job. AI scares artists like me because it normalizes any concept to the potential for virality and drops the technical barrier to near zero — which is exactly why the remaining differentiator is authorship. Originality matters less than authenticity. Cream still rises.
Where it goes
The workflow runs straight through to release and direct support: Bandcamp, a self-hosted Faircamp catalog on Cloudflare R2, and the economics I mapped out in the Purchase Power Calculator.