Caroline Hyde

And so this is a demonstration, we believe,

where different vendors at different times will be able to exhibit share gains with functionality that they launch in their ecosystems to enable value creation.

So we view it less as a zero sum and more of as an infinite game that we're just starting to play.

Alex, I find this so interesting that we've started this conversation about top-line growth in Azure, right?

Almost the reverse of a conversation we might have had one or two quarters ago, which would have started with capital expenditures.

How much do you need to just get a projection of confidence from Microsoft on their capital expenditure figures to keep this big story alive?

Yeah, so look, last quarter was the big shift quarter in terms of CapEx going from, you know, something where we thought it was going to be $100 billion to $150 billion this year.

We think that the CapEx shift has been made.

We think now it's a question of stability and, you know, to some extent, maybe aligning the increase from here.

with some of your kind of core revenue growth assumptions.

But look, that could change too.

Over the last month, we think inference has quite frankly exploded with the amount of tokens being generated for autonomous coding tasks that are now running, you know, writ large when people sleep.

as well as some of the possible functionalities exhibited by things like Clogbot and these other kind of orchestration engines.

So it's anyone's guess.

We don't think CapEx is going down anytime soon.

We think kind of stable to up is how we are thinking about it.

When it comes to genuinely autonomous agents, Microsoft's very good, I always say this in the program, it's what I grew up with, selling software that you can otherwise get for free elsewhere.

They're not the only game when it comes to the coworker.

AWS has a very similar story they like to tell.

They like to talk up their prowess in observability, a really important critical factor when an agent is an autonomous.