Claire Vo on AI and the PM Role

Claire Vo on AI and the PM Role

transcript product-management ai-product career cpto lenny-podcast

Claire Vo on AI and the PM Role

Guest: Claire Vo — three-time chief product officer; co-founder of ChatPRD; engineer, designer, and founder.
Host: Lenny Rachitsky
Source: Lenny’s Podcast, (earlier episode, pre-OpenClaw).


Overview

A broad conversation covering PM career growth, building fast-paced organisations, the CPTO (Chief Product and Technology Officer) role, being a woman in tech, and Claire’s views on how AI will reshape the PM function. The AI thread runs through the second half: ChatPRD as her personal experiment in understanding what AI can and cannot do in product work, and her framework for which PM skills AI will displace vs. amplify.


Key ideas

  1. Lowercase C vs. capital C communication. AI will automate functional information coordination (lowercase C — writing docs, sharing updates, synthesising status). It cannot replicate influence, boldness, and the capacity to move an organisation toward a vision (capital C).
  2. PM identity shift. The PM role will shift from “individual who manifests ideas” to “person responsible for whether the right things are built, built quickly, and deliver results — no matter the tool chain.”
  3. CPTO as response to AI-driven role collapse. When product, engineering, and design are increasingly done by the same people with AI tools, functional separation becomes overhead. The CPTO role unifies R&D under one accountable leader.
  4. Clock speed. Claire’s internal pace heuristic: “one click faster.” If you think something needs to be done this year → this half; this half → this quarter; this quarter → this month. Explicit acceleration, not just aspiration.
  5. Build non-deterministic product skills. PMs should practise building AI-native products (non-deterministic, probabilistic, fast-moving) as a career hedge — analogous to mobile specialisation in the 2010s.

AI’s impact on the PM role

Chip Huyen argues that talking to users is the highest-leverage PM activity; Claire Vo argues the same from a different angle: the functional communication and documentation work that occupies a large fraction of PM time is exactly what AI tools can automate.

Her taxonomy:

  • Lowercase C communication (automatable): PRD writing, status updates, synthesis, roadmap formatting, stakeholder briefs. ChatPRD does this.
  • Capital C communication (durable): getting a system of humans to follow you down a path, building trust, navigating ambiguity, making bold bets that are not in the training data.

Contrarian note on strategy: Claire agrees with Lenny’s observation that AI is particularly good at synthesis-based strategy work (given all known information, derive a plan). The durable strategic skill is seeing what is not in the training data — calling a future that nobody has written about yet.


CPTO role

Claire’s framing: product, engineering, and design are “one team building capital-P product.” The functional separation creates debates (what’s best for product vs. engineering vs. design) that distract from the real question: what is best for the organisation and the customer?

Two arguments for the CPTO role:

  1. Alignment. Unified R&D optimises for outcomes, not function.
  2. Accountability. R&D is the largest, most complex investment most companies make. One person responsible for that investment at the executive level improves decision speed and accountability.

Prerequisite: you have to be genuinely technical. The CPTO role is not “CPO with an engineering VP below.” It requires hands-on architectural understanding and operational engineering management.

See Agentic Engineering for how AI is accelerating the collapse of function boundaries (Boris Cherny’s observation that PM/EM/designer/finance all code on the Claude Code team).


ChatPRD

Claire’s AI PM tool, built as a personal experiment and released publicly. Started as a custom ChatGPT prompt, became a custom GPT, then a standalone app with the OpenAI Assistants API.

Key design: each user gets a personalised assistant that learns from their specific content, role, and company — not a one-size-fits-all GPT. ~60% of use is PRD generation from an idea, ~30% is PRD improvement, ~10% is brainstorming and internal PM work.

Qualitative user feedback: “saved dozens of hours on writing,” “extended our runway by not needing another PM.” The tool validates the lowercase-C automation thesis directly.


Clock speed framework

Claire’s pace-setting heuristic:

Natural tendencyClaire’s expectation
This yearThis half
This halfThis quarter
This quarterThis month
This monthThis week
This weekToday

Applied at the leadership team level. The mechanism: before stating a due date, check yourself and ask “is this right, or do I need to pull it in one iteration?” The anti-pattern it addresses: “let’s discuss this at the next meeting” — artificial cadences imposed by the calendar, not by actual timeline needs.


Career advice (PM-specific)

  • Know what you want and be explicit about it — with your manager, with your skip.
  • Frame promotion conversations around solving a problem for the company, not career growth for yourself.
  • PM is a generalist role; going “left and right” (marketing, operations, engineering) to go up is valid and often accelerating.
  • High-slope people get promoted as fast as the org can support. Do the work; the conversation follows.

See also