Historical music, contemporary genres, and a training corpus
A plan for turning historical music sources into structured essays, contemporary genre variations and a training corpus.
Lynn Cole / 标签
A thread through the work.
A plan for turning historical music sources into structured essays, contemporary genre variations and a training corpus.
Tell a coding system what good means: your architecture, standards, metaphors, file boundaries and expectations for documentation.
An early position on AI art and music: separate corporate training practices from artists’ work, and investigate the data before deciding what it demonstrates.
Building Grimes became an experiment in formal coding standards, logic maps and a self-bootstrapping agent workflow.
Music generation needs structural context: songwriting, orchestration and genre combinations grounded in more than a tag list.
Crunchy treats compositions as executable logic, with a two-way path between sheet music and code that can transform the performance.
A music-data flattener joins compositions, metadata and similarity records into structured training material—with a few late-night bugs still to fix.
A software product is a relationship. Understanding users, giving them a reason to care and making a living are part of the same work.
When a model calls a task chaotic, it may be recognizing a different kind of signal. I explore that response through Dada, metaphor and primordial chaos.
Preparing music-theory training data means organizing genres, generating explanatory material and checking the result before the first training run.
GhostTown’s podcast generator turns a social simulation into editable scripts, multi-voice audio and recurring stories people can follow.
Planning a comparison between a large Suno corpus and human-created music means solving data quality first and letting the results lead the argument.
The song-generator project needs music represented as machine logic. That requirement led me toward another notation language.
My predictions for 2026, written in December 2025: small models, edge compute, changing tooling, creative technology and the structures of power.
We have heard the vibe-coding argument before—in digital art, visual programming, web design, music and animation. Lowering the barrier changes who gets to try.
A generated voice can melt directly into an instrument. I enjoy the strange precision of that effect, even when it complicates a live-performance illusion.
Noise painting uses hand-crafted adversarial inputs and incongruent prompts to push diffusion into unexpected visual territory.
A dated perspective on AI work across software, commercial art, filmmaking and music, including my own job interviews.
Musical failure states, control settings, impossible instruments, and why I started building my own music pipelines.
A paranormal radio station, aliens who love pop music, and the music, creatures, sets and storytelling behind Night Signal.
Recursive poetry becomes a song, then a short film about evil robots and a software team that solves problems with meetings.
Taking Friday apart as a songwriting exercise: keep its narrative beats, rewrite the lyrics, and discover a different arrangement.
A Love Again remix became a new song through lyric writing, inverted prompts, vocal recasting and repeated reconstruction across music tools.
A keyboard sketch becomes a series of Udio arrangement experiments, with changes to similarity, source audio, genre and noise along the way.
Rewriting Joe Turner for a contemporary audience meant finding its structure, translating the details and carrying the emotional argument across genres.
I woke up this morning feeling this insatiable need to write a needy love song.
Invented words meet dance music in an experiment with Dada sound poetry, dissonance and the echoes that appear even in a made-up language.