Yeah, no, great

... yeah, the, I, I, I think it's really a function of what the market's perceived ROIC is on each of them on this capital build-out, right? And so, so let's take, uh, Amazon first, 'cause that was the clearest articulation by any of the hyperscalers on what the CapEx really would translate to into, to revenue and cash flow out of, out of Jassy.

So one, they, they broke out the different pieces of the CapEx, right? There's, there's long lifecycle assets and short lifecycle assets, right? You know, long lifecycle assets, you know, being property and, and, you know, actual infrastructure, and then the short lifecycle assets being, you know, the chips, the networking equipment, the da- the data center spend, right? Uh, and, and Jassy laid out on the call that on that short lifecycle piece, which is the piece that everyone would naturally worry about the most, that they break even in roughly two to three years, and then they last five to six years of useful life, right? So two to three years to get your money back, and then two to three years of, of great cash flow.

Um, and when you, when you think about what that means, it means that even on the shortest lifecycle piece of this, the company should have very, very strong ROICs, right? And then when you take that in the context of AWS growth accelerating to 37% at 50% incremental operating margins, you don't just have management saying ROICs are gonna be strong.

You see that coming through in the acceleration of the revenue growth and the EBIT margin expansion as well.

And then you take that in context of Microsoft Azure growing 43%, guiding to accelerating to 45% into context of Google Cloud Platform in to GCP growing 82%, also at 50% incremental operating margins.

I think you have a lot of evidence that the hyperscale CapEx business is hitting that inflection point.

How do you feel like that plays into this thesis, um, you know, not just with the hyperscalers in general, but the whole AI complex all the way down to, you know, memory and, and chips?

What gives you the confidence they can do a 10x on their top line in that time frame?

Well, let's see what happens, right? Elon, I think, has a tendency to have Brando statements and eventually achieve them, but maybe a few years out.

But regardless, whether or not he ends up hitting these exact targets, when I look at the street numbers, they look much lower than the potential forecast.

And you can look out a few years and have a stock that's trading at just a mid-teens earnings multiple, you know, right? So what pieces am I most confident? I think Elon's probably one of the best in the world at setting up AI infrastructure.

Jensen himself has said that.

And I think this exclusivity deal has really allowed him to get premium preferred compute from Nvidia.

And that's a unique place to be.

I mean, is there just enough for everyone to funnel that much capital and spending in that direction and not essentially, you know, really impair big parts of tech? Or is it somehow all additive?

As we are transitioning from chatbots to agents, what does that mean for this story and the pace and scale of AI adoption?

Well, let's first take a step back, Jennifer.

It's amazing to think about all that's changed over the past few years.

But what did I mean when we said AI has the potential to be the biggest productivity enhancer since electricity? I think AI is going to result in an unlock of productivity that's going to result in a strong economic boom.

And there's only three ways to grow an economy.

There's capital.

Throw more money at the problem, labor.

And that's why I think the landscape is going to look a lot more interesting and lively in the next few years with both the incumbents trying to develop AI agents, leveraging their existing relationships with the clients, and also the new entrants who want to take their super app and turn that into a consumer-facing agent interface.

Are you worried at all about concentration?

Of course.

So I think about it a lot.

So the first thing that I think is really important is not all semis are the same.

They all have different cycles, right? So we have analog semis, names like Texas Instruments, STMicro, Infineon, analog devices.

These companies sell into the auto and industrial and markets, right? And auto and industrial end markets have different cycles than the smartphone end market or the PC end market or the AI supercomputer market.

Those are very different end markets.

What did you learn from that? Big win or big loss?

There are a lot of familiar tech names, and I'm curious how you think about the balance between owning those big tech names and a- as you were mentioning, some of these, uh, th- these are magnificent companies, but how, how do you balance holding those while also finding some hidden gems, so to speak?

Yeah.

I'm gonna, I'm gonna separate the question into kinda two different pieces if that's okay.

So number one is Mag Seven and how I think about Mag Seven, and then I'm gonna go broader semi, and then I'm gonna go into some of these smaller ones.

So, so I'm doing a little bit of a weave, but I promise I'm gonna get there.

[laughs] Uh, on Mag Seven, okay? Look, I, I really believe you have to own an appropriate amount of Mag Seven.

These names have really strong downside capture.

So just generally speaking, how are you thinking about valuations right now?

Is it peak growth or is it competition? To your point that you just came back from South Korea, SK Hynix, I believe, uh, supplies more than half of Nvidia's high bandwidth memory that it needs for Rubin.

Yeah.

Well, I actually think Samsung's gonna do really well in this upcoming chip generation for high bandwidth memory.

But you gotta remember there's a few different things.

So one, there's the high bandwidth memory, which is actually lower margin now than core DRAM, right? DRAM demand has been so strong, and so core DRAM margins are actually higher than high bandwidth memory for the industry, something that you would have never thought, you know, just 18 months ago.

And I think what the market is struggling with is not necessarily competition, 'cause it's hard for all three major suppliers to bring on meaningful supply, but actually just the gross margins, right? The Micron number, 81% gross margins, that's a tough number to get much better than, right? So how are we gonna beat from these high levels? Now, all that being said, I still really like the AI setup here.

AI chips are going from $45 billion in 2023 to upwards of a trillion dollars by 2030.

In other words, are we adding as much capability for every dollar that the companies have to spend?