A field guide to the ML art zoo
Someone finally catalogued the chaos of neural nets making music, paintings, and fake Kanye raps.

What it does
This is a curated awesome-list that rounds up people, projects, articles, and tools at the intersection of machine learning and creative work. Think of it as a map to a scattered scene: Magenta’s music experiments, Robbie Barrat’s GAN paintings, GPT-3 fiction, and browser-based style transfer demos all live here. The maintainer, Vibert Thio, also slips in a few of his own musical ML games.
The interesting bit
The list doubles as a who’s-who of the creative ML world circa 2018–2019, including the then-19-year-old Barrat and the notorious Christie’s GAN auction that sold for $432,500—while noting the code was originally Barrat’s. It’s part directory, part mild art-world gossip.
Key highlights
- Covers visual art, music, text, and interactive installations in one list
- Includes learning resources tiered from beginner (The Coding Train) to advanced (CS231n)
- Highlights approachable libraries: ml5.js and p5.js for artists who fear C++
- Curated by someone actually building in the space (Sornting, RUNN, Jazz RNN)
- Creative Commons licensed, so the curation itself is reusable
Caveats
- The TODO section admits it needs linting, a profile picture, and a “For Non-Programmers” section—so it’s unfinished
- Some links may be stale; the Christie’s auction and GPT-3 references date the list to a specific era
- A few descriptions are thin (“Drawing Orientations” gets no explanation at all)
Verdict
Worth bookmarking if you’re exploring creative ML and want a human-filtered starting point. Skip it if you need rigorous taxonomy or up-to-the-minute tooling; this is a snapshot, not a living database.