Meta’s Muse: How we launched the most successful consumer product this year | Mona Sarantakos
Nov 15, 2026 · 1h 27m
More Technology podcasts · More Business podcasts
Summary
Mona, Head of Product at Meta Super Intelligence Labs, discusses the origin and development of Muse, the company’s new personal AI assistant. She details how the team prioritized safety and trust over speed, resulting in a seven-month launch delay to ensure reliability. The episode also covers the creation of the character Jolly, the surprising success of shopping use cases, and the unique, collaborative leadership dynamic with Mark Zuckerberg and Nat Friedman.
Topics discussed
Introduction: Mona Guo and the Muse product
Background: From Meta AI to Superintelligence Labs
The OpenClaw moment and the pitch for Muse
Challenges: Safety, security, and scalability
Why the seven-month development timeline?
Security vectors and browser integration issues
Explaining the concept to the public
Roles of Mark Zuckerberg, Nat, and Alex Wang
Defining Muse: 'Dreaming' and personal priorities
The Jolly avatar debate and design choices
Product principles: User agency and guardrails
Internal feedback loops and dogfooding
Surprising use cases: Shopping vs. Home Automation
How agents change retail, ads, and marketing
Using AI swarms for product insight and empathy
The evaluation framework: Evals and behavioral specs
Activation: Onboarding and the 'aha' moment
Key design decision: Main chat vs. multiple threads
The 'Let them cook' culture and engineer autonomy
Fun story: Two Muses talking and fighting
How AI changes the PM workflow and prototyping
The new bar for PMs: Taste over execution
Design as a differentiator and the board meeting story
Team structure: Keeping the core team small
Building for ourselves: The most personal product
User stories: Parents and busy professionals
Democratization and the fear of 'babysitting' agents
Future vision: Health, time, and feedback culture
More from Lenny's Podcast: Product | Career | Growth
Listen ad-free on Castria