What We Took Away From Brian Chesky on Rebuilding Airbnb for the AI Era

Most conversations about AI start with the product.

What gets automated?

What gets personalized?

Which workflows become faster?

Brian Chesky’s recent conversation with Patrick O’Shaughnessy goes somewhere more interesting.

He asks what happens when AI changes the organization itself.

Chesky describes what he calls “AI Founder Mode”: fewer layers, more direct access to information, less dependence on meeting-heavy management, and managers who remain close to the actual work.

For operators, that is worth thinking about.

Management layers are often information layers

As companies grow, founders rarely become less capable of making decisions.

They become further away from the information required to make them.

One manager becomes three.

Three becomes ten.

Frontline reality moves through several layers before reaching leadership.

By the time the information arrives, it has often been simplified, filtered, or translated.

Chesky has spoken openly about experiencing this at Airbnb before the pandemic, when the company had grown to thousands of employees and he felt increasingly disconnected from what the organization was actually doing.

AI potentially changes that.

Not because it replaces management.

But because access to information can become much more direct.

The implication is not “remove managers”

This is where we would be careful.

Flattening an organization does not mean management stops mattering.

In physical consumer businesses, management may matter even more.

A store manager still needs to coach people.

An area manager still needs to diagnose performance.

A warehouse leader still needs to make judgment calls.

But pure information-passing roles become harder to defend.

If a manager’s value comes mainly from collecting updates from one layer and forwarding them to another, AI can compress that role.

The stronger manager becomes someone who:

understands the work,

has judgment,

can diagnose problems,

develops people,

and improves execution.

AI may reduce the value of managing information while increasing the value of managing the work.

This matters for retail too

Imagine a 500-store retailer.

Historically, headquarters may only see store performance through dashboards and regional reporting.

AI can potentially make thousands of operational signals more accessible:

customer complaints,

inventory anomalies,

labour productivity,

stock-outs,

training gaps,

store-level sales patterns,

maintenance issues.

The question then becomes less:

“How do we build another reporting layer?”

And more:

“How do we make better decisions with fewer layers between reality and leadership?”

That is a very different organizational problem.

Our Conviction

AI will create better consumer products.

But one of its less discussed impacts may be organizational design.

Companies may become flatter.

Managers may need deeper functional expertise.

Information may travel faster.

And leadership may be able to remain closer to the actual work even as the company becomes much larger.

The opportunity is not to eliminate management.

It is to eliminate management that exists mainly because information used to be difficult to access.

The companies that benefit most from AI may not simply have better tools. They may build fundamentally better operating systems around them.