Monologue: What The Hell Are The Hyperscalers Doing?

0:02 Zone, media, superstition, fear, and jealousy. Welcome to this week's Better Offline monologue. I'm your host ed ze Tron. As ever, please subscribe to the newsletter. We're running a one year anniversary deal right now. The link will be in the episode notes, and we're giving you ten bucks off the annual rate for life and even if you die and come back to life, I'll give it to you anyway.

0:34 The deal runs through midnight on Sunday. But if you happen to hear this after this point, email me at easy at Better offline dot com or ez at better Offline dot com for the British and Canadians in the audience, and I'll sort you out. I'm generous like that, and I love my premium mates. Really some of the best analysts work out there. It's very deep. I got the terminal going now and pulling all sorts of data. The many numbers I read every week are driving me truly insane.

0:57 And with that in mind, to day's monologue starts with a simple question, what the fuck are hyperscaler is doing? Now? I realize I've asked this question maybe one hundred thousand times in the last two years, but it's begun to get so incredibly stupid that I'm actually a little exasperated. McKinsey.

1:15 Spawn and Google CEO Sundar Pashai announced in Wednesday's earnings that Google would raise its year end capital expenditures to somewhere between one hundred and ninety five billion dollars and two hundred and five billion dollars after burning eighty point six billion dollars in the first half of the year, and this suggests that Google will spend another one hundred and fourteen billion dollars by the end of twenty twenty six.

1:35 They're going to be doing what fifty five sixty billion dollars a quarter. It's fucking insane, it's fucking stupid. I'm sorry for swearing, but it's absolutely deranged. What are you doing, Sunda. You've already spent two hundred and eighty eight billion dollars in capex since twenty twenty two, and all you've got to show for it is several also RAN llms and one indeterminately large customer in Anthropic, who you've also invested thirteen billion dollars in. Now. One answer might be that Anthropic helped choose Google cloud's growth by eighty two percent year of a year in this qure But considering that Google doesn't actually break out how much of that is from AI, or from Anthropic, or really from anything, it's

2:14 difficult to say what exactly this money is doing, other than turning its free cash flow negative for the first time since Google went public. While I can't say for sure how much Anthropic is spending, I do believe it's taking up the vast majority of Google's AI infrastructure and representing most of its cloud growth. Google Cloud was, up until fairly recently the last year or so, are much smaller business than Microsoft. As euroamason web services have been growing somewhere between twenty two percent and thirty percent year of a year, hitting thirty three point nine percent year of a year growth in the third quarter of twenty twenty five. That very same quarter, Google and Anthropic announced

2:49 the deal to expand its use of TPUs, which is Google's special GPUs, in a contract worth than I quote tens of billions of dollars. Mysteriously can't imagine how Google's next time three quarters of cloud revenue growth showed forty seven point seven percent sixty three point four percent and eighty two point one percent year of a year growth.

3:07 Whatever could that be? Where did that come from? I can't work it out, can you, dear listener? What could it possively? It's anthropic, It's it's obviously anthropic. Except in those same quarters, Google's capital expenditures went up ninety five percent, one hundred and seven percent, and one hundred percent year of a year, absorbing every dollar of cloud revenue and then some. In fact, outside of the fourth quarter of twenty twenty four, Google's capital expenditures have outpaced cloud growth by at least twelve percent year of a year for two goddamn years. And if your pushback is, well, Google spending that CAPEX on something else, No, they're not. They're not.

3:44 Maybe they've got a couple of billy coming in from Google Workspace and cucking their various subscribers with annoying things. It's coming from anthropic. Google is not making that money from Gemini. Your ass must be crazy if you think that's that's the point. But all of this begs the question, when exactly does all of this payoff? How does it payoff?

4:06 What does the payoff even look like? And why am I one of the few people asking these questions, hyperscalas have now spent I asked them at about one point one trillion dollars. I haven't seen Amazon and Microsoft's quarters yet, and I'm going to guess they spend a good amount on capex. They now spent one point one trillion bloody dollars on capital expenditures. And when you ask them to tell you how much revenue they made from said expenditures, they look at you like that meme of Brendan Fraser from the whale. Anyway, here's a fun fact for you.

4:35 Thanks to my own reporting and open ais auditive financials, we know that It's spent seventeen point two billion dollars on Microsoft a zure in colendar year twenty twenty five, a year when Microsoft's intelligent cloud segment made one hundred and twenty point four billion dollars in revenue. The segment grew as a result twenty six percent year of a year. Yeah, I'm a curious little Chris or around the numbers, and open Ay's revenue represented fourteen point three percent of all Microsoft's cloud revenues for that period. And to be clear, it's calendar year twenty twenty five, they have a fiscal core thing. But I went and worked it out. But this means the open Aiy's money represented eighteen percent of

5:13 the year of year growth of Microsoft's cloud. Without it, their cloud revenue growth was a mere eight percent year of a year. That's pathetic. That's a bonable. That's barely keeping up with inflation. That's dog shit. Satya, Satchy, you should be ashamed of yourself. Sacha, come on my show so I can ask you very reasonable questions. I'll be nice.

5:32 I swear you won't. You're scared, just let the rest of them, just like Clammy Sam on one Clamuel, your public relations people said you're too scared to speak to me. Well, they said you wouldn't speak to me, and I know you're scared. Same with weird Daria. Where are you boys? You're too scared of better offline? Anyway, Enough of the Trump voice, though, there is actually a reason that hyperscalers are still spending all that capex, open AI and Anthropic allow hyperscalers to boost growth by feeding their own money back into their businesses, a short term solution to the long term problem of the end of the era of hypergrowth.

6:06 One of the reasons they're so shy about AI revenues is that they know it'll become obvious where those revenues are coming from two unprofitable and unsustainable AI labs and a little trickle, little spurt from AI services they dump on their enterprise customers through confusing subscriptions up cells and just adding AI to random things and being like, look, folks, don't we love this? Don't you all love this? You little shits? And hyperscalers also know that the second they cut their capex spend, their AI trade unwinds. If Microsoft, Google, Amazon and Meta aren't buying hundreds of billions of dollars of GPUs a year, that means Taiwanese server manufacturers like Quanta and on High won't sell as many servers and

6:51 thus won't need as many GPUs broad common in video won't need as much high bandwid memory for those TPUs and GPUs, and they are sent speculators across the world will suddenly realize that they conflated hyperscalers being stupid with their money with the existence of massive amounts of demand for AI compute. But another way, hyperscalers are doing k fape to keep the crowd riled up, but everybody slowly working out that the wrestlers aren't really hitting each other and everything is scripted. I also should add I can just anticipate I'm going to get an email from someone being like, well, actually, this is good. The cattle expenditures are good because it showed growth, even though it's short term.

7:26 That's a silly, silly perspective to take, considering they're feeding all the revenues back into the business in the capex for stuff that won't be useful in the future. Mathematics. My friend try it sometime. Anyway. It is a funny old time for me right now, because I feel like I'm on three or four podcasts or TV shows or interviews a week, as the system appears to be trying to cover its ass by having me on to explain stuff I said two years ago, but a little more insistently and clearly than before. While the mainstream media seems hesitant to fully commit to the graveyard small of the AI bubble. It feels as if there's a conscious attempt to represent my work as more than just the other side,

8:05 or a hater or a skeptic or a cynic, which is both edifying and enraging because we are now so far past past the point of no Returnemy. Yeah, we really can't avoid what's coming. As I wrote in my newsletter this week, I think there's a subprime data center crisis brewing right now, with over five hundred billion dollars in data center SPVs that can only ever make investors whole if there's hundreds of billions of dollars of AI compute demand in the next two years. And to be clear, when I say hundreds of billions of dollars demand, I mean a year. I mean literally a year. I ran the numbers, and for the what I think it's one hundred ninety giga wats of supposedly planned data centers, you'd need one point six trillion dollars a year in annual

8:47 compute revenue to make it worth worthwhile. And by the way, the global software market is under eight hundred billion dollars a year, So I don't know where that fucking money's coming from, and no one else seems to really want to think about him. And I deeply worry about what happens when the demand doesn't arrive. As these data center debt deals are often funded by private credit funds, which are in turn funded by both public and private pensions and insurance funds. Anyway, do read the free newsletter to learn more. It is free this one. I have a premium one coming out today that will be about well, the fact that Oracle's dying, which I mean, hey, you in the back, stop laughing. Stop it, poor Larry Ellison

9:26 Paul who gives a fuck in all seriousness. Though, it's going to be a banger. The Free ones a banger. There's bangers everywhere. And I'll be back next week with my good friend Nick Saresh to talk about how AI mania is eviscerating global decision making, creating a cult lug environment where once participation in AI is almost a religious act, and if you talk badly about AI, it's considered a cardinal sin or a cardinal sin, one of those two.

9:50 Sure one of you will correct me on this one. Nevertheless, it's a really he it's based on this incredible piece he wrote. I can't wait for you to read it. I can't wait for you Timmy talk to him. Oh no, can't wait for you to consume my content. What a peculiar thing to say. Anyway, I'll got you Next week Zitron out