Descriptions:
To open 2026, Peter Yang hosts fellow AI product managers TA and Aman for a live demo session where each shares the AI workflows they actually rely on day-to-day — not theoretical setups, but the specific tools and habits that have changed how they work as PMs.
Yang leads with his Claude Projects setup for writing strategy memos: he uploads a one-page template and research notes, then collaborates with Claude to draft a Google AI Studio strategy memo in real time. The group critiques the output together, illustrating how they use AI as a thinking partner rather than a one-shot generator. TA demos rapid product prototyping — he built and user-tested a working prototype before his designer had finished the Figma mocks, turning AI-generated code into a feedback loop that runs ahead of traditional design timelines. Aman shows his personal OS built on Obsidian: markdown files that act as an evolving personal context layer, with Claude reading and editing them as a long-term accountability partner and focus tool.
The most practically useful segment is an extended discussion of Cursor versus Claude Code as daily drivers. TA and Yang converge on a clear mental model: Cursor feels like pair programming — you see diffs, approve changes, stay in the loop — while Claude Code is better described as delegation, where it goes off, does the work, and reports back. For strategy and thinking work where staying in context matters, they prefer Cursor. The conversation includes a spontaneous live revert when Claude Code’s changes go in the wrong direction, making the tradeoff concrete rather than abstract.
📺 Source: Peter Yang · Published January 11, 2026
🏷️ Format: Showcase
