Simon Willison's blog offers a mix of technical tutorials, data analysis projects, and reflections on technology and society. With a focus on Python, SQL, and web development, Simon shares insights into building web applications, working with data, and exploring emerging technologies. Developers can learn about data visualization, API design, and software engineering best practices, gaining inspiration and practical knowledge to advance their careers.

Simon Willison

FastRender is a web browser rendering engine built from scratch using thousands of parallel AI coding agents coordinated by Cursor. Started as a personal experiment in November, the project ran up to 2,000 agents concurrently, generating nearly 30,000 commits over three weeks. The agents worked autonomously for up to a week at a time, organized in tree structures with planning agents dividing work and worker agents executing tasks. Key insights include: agents successfully minimized merge conflicts through effective task division, GPT-5.1/5.2 outperformed specialized coding models for autonomous operation, and allowing temporary compilation errors proved beneficial for throughput. The project demonstrates what a single engineer can achieve with agent swarms in 2026, producing over a million lines of Rust code capable of rendering real web pages.

Wilson Lin on FastRender: a browser built by thousands of parallel agents