Lenny's PodcastJulie Zhuo: How AI is flattening orgs into broader builders
At Sundial, no dedicated PMs: engineers cover design and analysis; Zhuo diagnoses with data and treats with design, leading like a sturdy, flexible willow.
At a glance
WHAT IT’S REALLY ABOUT
Managing AI Like People: Julie Zhuo’s Blueprint for Future Leaders
- Julie Zhuo argues that the core skills of great management—setting clear goals, understanding strengths, and designing effective processes—map directly onto managing AI models and agents.
- She describes how AI is flattening organizations, shrinking team sizes, dissolving rigid roles, and turning everyone into broader “builders” who lean on AI to do more of the work themselves.
- Drawing on her journey from Meta’s head of design to founder of AI analytics startup Sundial, Julie explains how to “diagnose with data and treat with design,” and why most fast-growing AI companies are still underusing data.
- She also shares timeless management advice on self-awareness, feedback, emotional regulation, and leading through rapid change, framing modern leadership as being “sturdy while flexible” like a willow tree.
IDEAS WORTH REMEMBERING
5 ideasManaging AI is still management: define outcomes, know strengths, design processes.
Julie frames AI as a new kind of resource: instead of only assembling teams of people with specific skills, leaders now also assemble models and tools with different strengths. The fundamentals—clear goals, understanding capabilities, and structuring effective workflows—are unchanged.
Org structures are flattening; everyone is becoming a broader “builder.”
AI lets individuals cover multiple traditional roles (design, coding, analysis) to a solid level, reducing the need for large, specialized teams. Julie’s startup intentionally avoids rigid role definitions and even often operates without PMs so engineers and others fully own product thinking.
Clarify success with ruthless precision—or your AI and your team will flail.
The hardest, most valuable skill for both managers and AI users is defining success in unambiguous terms. That clarity powers better prompts, better evals, better metrics, and ultimately better decisions, especially in conversational and agentic systems.
Use data to find the problem, not to dictate the solution.
Julie’s mantra is “diagnose with data, treat with design”: data (quantitative and qualitative) helps you understand reality, spot issues, and locate opportunities, but it cannot tell you what to build. Great teams pair rigorous diagnosis with creative solution design.
Fast-growing AI companies are often flying blind on data.
Because AI products can scale insanely fast with tiny teams, many hot companies lack mature logging, instrumentation, and observability. They ride intuition and hype until growth stalls, and only then scramble to build proper data foundations and analytics.
WORDS WORTH SAVING
5 quotesYou want to diagnose with data and treat with design.
— Julie Zhuo
It used to be people, but now it's basically models, and different models have different strengths. You kind of have to assemble the Avengers.
— Julie Zhuo
We need to dissolve the boundaries of these traditional roles and call ourselves builders.
— Julie Zhuo
Today, management is really about this idea of be sturdy while being flexible. I think about this metaphor a lot of the willow tree.
— Julie Zhuo
Every strength is its own weakness, and every weakness is a strength.
— Julie Zhuo
High quality AI-generated summary created from speaker-labeled transcript.
Get more out of YouTube videos.
High quality summaries for YouTube videos. Accurate transcripts to search & find moments. Powered by ChatGPT & Claude AI.
Add to Chrome