More agents, same human brain

This article explores the author's experience of running multiple AI coding agents (Claude Code, Codex, Gemini CLI) simultaneously for software development. While initially exciting, the author discovers the real challenge is not starting more agents but keeping up with them. The piece examines how AI changes the traditional engineering capacity model from scaling with people to scaling with agent orchestration. It discusses practical motivations like utilizing multiple AI subscriptions efficiently and highlights that generating more code doesn't equal creating better software. The article offers insights into the future of software engineering where human judgment, system understanding, and ownership remain crucial.