Nikesh Arora

You are seeing the consumer case.

I think we underestimate how much the entire consumer landscape is changing and is going to change in the next 24 months.

The idea of having applications where we have to go do all the work ourselves is going to become arcane.

You'll have agents that'll go book your Uber, will go get you your food from DoorDash, will book your airline ticket,

We can all imagine a future like that.

We all want a future like that.

That's going to require significant amounts of compute at the end.

So I think from an infrastructure perspective, in history, we've never had a situation where we built infrastructure, it didn't get consumed, and we wanted more.

So I don't think the infrastructure problem is a real problem.

Whether the demand comes right away or it comes an year or two later, I think that time will tell.

But from a demand, from a technology fit, we're there.

From an AI perspective, enterprises are slightly lagging that consumer adoption, but I think they're actually working hard to get there.

Every time we do some AI experimentation internally, we come back and say, wow, that is cool.

I wish we could deploy that faster across the entire enterprise.

I think the demand is going to come.

The timing is up to you guys.

You guys follow that on a daily basis.

We don't.

We put our heads down and see, where am I going to be three to five years from now?

How do I position my auto network in that context?