Mitchell Hashimoto describes how he went from AI skepticism to building a daily workflow around agentic AI tools. He forces himself through inefficiency, learns how to structure tasks so agents succeed, and gradually shifts from “AI as helper” to “AI as background worker.” Over time he develops harnesses—tools and prompts that prevent repeated mistakes—and aims to always have an agent doing useful work while he focuses on what he enjoys. The journey is measured, practical, and focused on craftsmanship rather than hype.
Catalog response pending