AI amplifies PM roles: Why product direction beats iteration speed
🎧 PodShort
59 min squeezed to 2
AI SprinklerAS AI / ML New

Sherif Mansour
Head of AI & Product Craft at Atlassian
Full episode from Lenny's Podcast
Quotable Moments
The narrative was everyone's becoming an AI builder.
We realized throughput's not our problem, direction's our problem. I need more people to steer a ship.
AI is really good at amplifying your strengths, and equally it amplifies your weaknesses.
Key Insights
- The initial narrative that AI would make PMs obsolete was incorrect; instead, AI is amplifying the need for strong product direction and context-setting.
- In a large organization, the biggest challenge for PMs isn't iteration speed (as AI improves throughput), but providing clear direction and filling context gaps for the team.
- The most effective PMs are those who focus on building a 'shared brain' within their team, empowering engineers and designers to make decisions by providing them with the necessary context.
- AI is not a substitute for domain expertise; having access to an AI tool doesn't make someone an expert in the problem space.
- The best PMs are not necessarily the ones who make all the decisions, but those who empower their team to make better decisions by providing context and facilitating collaboration.
- The rise of AI tools is making 'artist PMs'—those who lead with vision and intuition—more valuable, as they can better guide AI in creative problem-solving.
- The most impactful PMs are not those who possess every skill, but those who can identify their strengths, lean into them, and build a complementary team to cover their weaknesses.
- The trend is that PM roles are seeing the most growth in job postings, even as some fear AI might replace them.
Metrics Mentioned
- 100 open PM roles (In tech companies currently, compared to other roles like engineering or design.)
- Engineering roles seeing a slowdown in open postings (Contrast to the growth in PM roles, potentially due to AI's impact.)
- Design roles flat in open postings (Contrast to the growth in PM roles.)
- 10 million AI agents (The potential scale of AI agents needing to register, authenticate, and gain permissions, handled by WorkOS.)
RevBots.ai View:
- AI Sprinkler teams bolt AI onto workflows but miss the ARM opportunity: rethinking decision flows.
- PM-AI collaboration mirrors ARM principles: AI handles throughput, humans steer strategy.
- Context-setting PMs are the revenue team equivalent of ARM's orchestration layer.
- SaaS Hoarder orgs will misuse AI as a 'productivity tool' instead of rebuilding team dynamics.
Join The RevBots ARMy
The insider daily for Autonomous Revenue Masters.