Lenny's PodcastHow the most AI-pilled product team builds products | Fiona Fung (Claude Code and Cowork)
At a glance
WHAT IT’S REALLY ABOUT
Inside Anthropic’s AI-native product teams: velocity, verification, culture, and planning
- Anthropic teams are seeing massive throughput gains (e.g., ~8× more code shipped), shifting engineering constraints from writing code to verifying quality, safety, and user experience.
- AI-native teams blur traditional roles as engineers, PMs, and designers increasingly all “build,” with managers staying hands-on through dogfooding and shipping real PRs.
- Fiona uses Claude Code sessions, routines, and repo-integrated visibility to turn management into a data-and-artifact-driven practice (summaries, themes, and even draft PRs) rather than status meetings.
- Hiring priorities split into two profiles: creative product builders who iterate from user feedback and deep systems experts who can validate hard constraints and architecture in a trust-but-verify world.
- The next frontier is asynchronous work with fleets of agents, creating new challenges like context switching, loneliness, skill atrophy, and maintaining culture during rapid scaling.
IDEAS WORTH REMEMBERING
5 ideasCoding speed is no longer the constraint; verification is.
With AI-assisted commits and dramatically higher output, teams must invest in tests, monitoring, evals, and clear definitions of “good” so quality keeps pace with shipping velocity.
Put “what good looks like” into the repo to scale automated review.
Fiona argues Claude is strongest when it can review against explicit frameworks (specs, content guidelines, expectations) that are versioned alongside code and kept up to date.
Use AI to manage by artifacts, not by meetings.
She runs Claude Code sessions tied into repos and Slack to review what shipped, how it performed, and what themes emerged—often arriving with summaries and candidate PRs for discussion.
Asynchronous agent workflows will become the dominant operating model.
Routines act like cron jobs that monitor channels, synthesize feedback, and spawn work while humans sleep—pushing teams toward higher abstraction and more parallel execution.
High-agency teams require explicit accountability mechanisms.
Anthropic’s ethos is “freedom to cook,” paired with accountability via hypotheses, clear ownership, and verification loops so autonomy doesn’t turn into uncontrolled change.
WORDS WORTH SAVING
5 quotesCoding is no longer the bottleneck. It's lifted the ceiling of what anyone is able to do
— Fiona Fung
Make new mistakes. Like, it's okay to make mistakes, just make new ones so that we're always learning.
— Fiona Fung
For anything that there is a fear, my advice is lean in and ask, "What can I do about it? What is within my control?"
— Fiona Fung
We say with high agency is also high accountability, so it's all about making sure folks have that freedom to cook. But then it's also like, okay, what's the accountability for it? What's a hypothesis of what you're trying to solve?
— Fiona Fung
Explicit permission to kill processes that no longer serve us.
— Fiona Fung
High quality AI-generated summary created from speaker-labeled transcript.
