4 people outperform 10 in a committee.
That's the data that changes everything.
"A strike team of 4 beats a committee of 10."
This isn't metaphor. This is what real companies report.
What a16z discovered
a16z (Andreessen Horowitz) looked at the architecture of AI-first companies. The pattern is stark: smaller teams, faster code, fewer meetings.
But why?
Because when you build from scratch with AI as your operating system, you don't inherit the bureaucracy. You don't have legacy systems fighting new systems. You don't have 8 stakeholders who need consensus.
You have: one product. One velocity. One reason for existing.
The retrofit problem
Now compare that to companies trying to bolt AI onto existing structures.
They bring in a "Chief AI Officer." They create committees. They build AI centers of excellence.
And then they realize: the organization isn't designed for speed. The culture isn't designed for risk. The decision-making is designed for safety.
AI doesn't fit. The company has to break itself to make room.
If this plays out
Within 5 years, the market separates into two categories:
- Companies that built AI-first from the ground up
- Companies that tried to retrofit
The first group will move at 3x the speed. Their unit economics will be better. Their teams will actually want to work there (no committees).
The second group will still be in "pilot" mode, building committees, hiring consultants, waiting for the next quarterly review.
Who wins
The startup ecosystem. The small teams. The people willing to break everything and rebuild.
Who loses
The Fortune 500 that waited too long. The managers who thought they could slap AI on top of hierarchy. The HR departments that still measure productivity in "seat time."
Want more on managing in the AI era?
The Little Book for New Managers has helped 30,000+ managers figure out what actually works.
Get the book ($19) →