In this monologue podcast, the speaker enthusiastically introduces ClaudeCode hooks, emphasizing their importance for observability and control in agentic coding. The speaker details how these hooks—pre-tool use, post-tool use, notification, stop, and sub-agent stop—allow engineers to observe, iterate, and improve their agentic systems. The speaker also discusses the recent move of ClaudeCode's creators to Cursor, speculating on the reasons behind this shift, including compensation, creative control, and defensibility of ClaudeCode. The podcast further promotes aicoding principles and a related course, highlighting the value of talent and information in the generative AI age, and positions ClaudeCode as a key tool for increasing compute advantage and scaling AI coding efforts.
Sign in to continue reading, translating and more.
Open full episode in Podwise