How Anthropic’s product team moves faster than anyone else

Cat Wu, Head of Product for Claude Code at Anthropic, discusses how her team ships at unusually high speed, how the PM role is shifting toward product taste and fast iteration, and how Anthropic's mission shapes its decisions. The conversation also covers Cowork, AI-driven workflows, evals, and the skills PMs need as models improve.

01Key learnings

• Shorten the path from idea to user: AI-era timelines have shrunk from six months to days or a week, so PMs should focus on fastest-to-ship paths over multi-quarter alignment.
• Ship features as clearly labeled research previews to lower the commitment of launching and get fast user feedback that can be iterated on next release.
• Set a clear, specific goal for who the key user is and what success means, which rules out many approaches and speeds decisions.
• Build a repeatable launch process with cross-functional partners (docs, marketing, DevRel) pre-agreed so engineers can ship with minimal friction.
• As code gets cheap, product taste, deciding what to build and how it should feel, becomes the most valuable PM skill.
• Revisit prompts and features whenever a new model ships; remove crutches the model no longer needs and build prototypes of features the next model may enable.
• Automate repetitive work with AI, but push automations to near 100% reliability through feedback, since a 95% automation often isn't worth using.
• Build a small set of evals and use trusted power users for fast qualitative model feedback, since evals help quantify goals and progress.

02Frameworks mentioned

Summary and takeaways written by PM Atlas; quotes are short excerpts. © the original author.