Felix Rieseberg, the Anthropic engineer who built Claude Cowork, sat down with Lenny's Newsletter to explain exactly how he uses the tool he made. That framing matters: this is not a product demo. It is a practitioner describing his own daily workflow, including where the tool works, where it fails, and why he built it the way he did.
The conversation covers concrete usage patterns: how Rieseberg structures tasks for Claude, how he thinks about context management, and what Claude Cowork does differently from a standard API call or claude.ai session. The original is worth reading for the specific prompting and orchestration details, not just the high-level philosophy.
This comes out alongside Google I/O 2025 coverage, which makes the timing pointed. Two competing visions of how engineers actually integrate AI into daily work are on the table at once. The Rieseberg interview is the more grounded of the two.
[READ ORIGINAL →]