Overview

Cursor’s research team built a functional web browser from scratch using thousands of autonomous AI agents working in parallel. The project demonstrates AI agents can now coordinate on complex, long-term software development tasks without human oversight.

Key Facts

  • Started as Wilson Lin’s personal experiment with frontier models (Claude Opus 4.5, GPT-5.1, GPT-5.2) - AI can now tackle extremely ambitious engineering projects independently
  • Successfully renders GitHub, Wikipedia, and CNN pages despite missing JavaScript engine - agents can build working software even when incomplete
  • Agents made autonomous decisions like disabling JavaScript via feature flags - AI now makes architectural decisions without human input
  • Graduated from side project to official Cursor research when single agents showed promising results - companies are betting on multi-agent development as the next breakthrough
  • Browser rendering engine chosen because it’s complex but well-specified with visual feedback - agents work best on tasks with clear success criteria
  • Goal was never production software but observing multi-agent behaviors at scale - this is research into AI’s future capabilities, not a Chrome competitor

Why It Matters

This represents a major leap in AI autonomy - agents can now coordinate on months-long software projects. It signals we’re moving from AI as a coding assistant to AI as an independent development team.