William Mapan on blurring code, paint, and generative art (www.artbasel.com)

French artist William Mapan is debuting a new hybrid body of work at Art Basel Paris: a run of 250 one‑of‑a‑kind generative prints (each 1/1) alongside related paintings in the "Code + Matter" show. Mapan — a former creative developer who studied animation and software engineering — blends hand‑made paint textures, landscape photographs, 3D models and algorithmic composition. He iterates prolifically (100–500 code sketches a day), sometimes using a plotter to render drawings, and usually codes in JavaScript with a bespoke toolbox rather than relying on open‑source libraries. The project grew out of pandemic reflection and the NFT era that made digital work commercially viable. For the AI/ML community this practice is a useful case study: generative art thrives on tight control of the algorithmic pipeline, material inputs (scanned paint, photos) and rapid iteration, rather than black‑box automation. Mapan welcomes AI for repetitive tasks (exporting, file management) but warns against outsourcing creative coding to opaque models — doing so risks losing the affordances and discoveries that come from direct tinkering with code. The work underscores demand for web‑friendly, transparent tools that integrate with artists' workflows and suggests ML research should prioritize controllability, provenance, and tooling that augments rather than replaces artist-driven algorithmic exploration.