Sketching with code in the age of AI

by Qianqian Ye and John Maeda

In this GitHub Universe 2026 session, Qianqian Ye examines creative coding as a way to understand and shape what AI produces. John Maeda joins her to discuss the connections between design, computation, open source, and creative judgment.

Understanding code beyond syntax

AI tools can generate clean, functional output from a short prompt. The session shifts attention from producing that output to understanding what has been made.

Drawing on her experience teaching creative coding at the University of Southern California, Ye discusses sketching with code using p5.js and working with GitHub Copilot. Her focus is on how experimentation and iteration develop the intuition needed to shape systems rather than only prompt them—and what is lost when that process disappears.

Slowing down, noticing, and creative judgment

Ye connects the discussion to her experience navigating AI as a new mother and her “becoming a mother” project. The session explores why slowing down matters when working with AI, alongside the role of play and noticing in creative practice.

The conversation also addresses the legacy of Processing and p5.js. Maeda joins to explore open source longevity, the role of open source in the age of AI, and the tension between AI-generated output and creativity. Together, they consider the creative judgment that AI cannot replace and their hopes for the next generation.

Session chapters