00:04:18 · John Coogan, Jordi Hays
[00:04:18] S1: Interesting. So, yeah. I wonder what that relationship's like. I mean, OpenAI has, you know, I I I felt like they were always be there would always be some relationship with Microsoft because Bing has really solid search infrastructure. Like, for a while, people were saying like, Bing has as robust as a of a crawling infrastructure as Google. Like they built the technology. They just weren't able to actually transform, you know, customer behavior to the degree that people would by default go to Bing. They ran this big campaign, Bing it and Go and all this. And it never really happened, but the tech was always good because they were they knew what to build and they went and built it. They just couldn't change people's behaviors. So interesting. I wonder if Google will be bringing out the sharp help us anytime soon for OpenAI. Anyway, let me talk about ramp. Time is money. Save both. Easy as corporate cards, bill payments, accounting, and a whole lot more all in one place. Go to ramp.com. So Tane, who's been on the show before, says Sundar Pichai on Google CapEx. With this strong and growing demand for our cloud products services. We are increasing our investment in capital expenditures in 2025 to approximately 85,000,000,000 and are excited by the opportunity ahead. We gotta hit the gong. [00:05:28] S0: Hit the gong for Sundar. Well deserved. Sundar pitch AI. [00:05:34] S1: He's pitching AI. He's pitching it successfully. The the the Google earnings are staggering. So 14% jump in year over year revenue. Guess their ARR, their annualized revenue.
00:05:34 · John Coogan, Jordi Hays
[00:05:34] S1: He's pitching AI. He's pitching it successfully. The the the Google earnings are staggering. So 14% jump in year over year revenue. Guess their ARR, their annualized revenue. [00:05:49] S0: Wow. I don't even oh, run [00:05:51] S1: Their run rating, 400,000,000,000 almost. [00:05:54] S0: Okay. [00:05:54] S1: They did they put up 96,400,000,000.0 in the second quarter and said CapEx. It's like, it would be a lot if it was in millions. It's in billions. They said that CapEx would increase 13% to 85,000,000,000, and that compares with 52,000,000,000 last year. So that's a big, big increase even above expectations. Ben Thompson has a great analysis of this and is it came away, from my reading, pretty excited about it. So Andrew Curran also has a post here. Says, Sundar Pichai on the earnings call, the growth in usage has been incredible. At IO in May, we announced that we processed 480,000,000,000,000 monthly tokens across our surfaces. So that's both AI search features and Google search AI mode, Gemini, also API stuff. So overall, tokens, we're gonna have Micky Malka on the show. Later. He's gonna talk about token factories. We're The demand for tokens is is immense. And and it's and it's really just weaving its way into all these different nooks and crannies. It's certainly going mainstream. If you just ask a friend who's not, you know, in the horse race of who's gonna win just a just a normal friend, look at let's look at your screen time. How much time are you looking on ChatGPT? You're gonna see ten minutes. [00:07:09] S0: Gonna see thirty
00:07:09 · Jordi Hays, John Coogan
[00:07:09] S0: Gonna see thirty [00:07:10] S1: minutes a day. [00:07:10] S0: This was buried, not in the main post. But Andrew follows up and said, he said there would be over 70,000,000 user videos made with v o three which is which is impressive especially considering that they massively rate limit everyone. Yeah. You you pay $500 a month and you get a measly couple videos a day. [00:07:27] S1: It's hard. It's hard to [00:07:28] S0: actually Somebody made a great very cool intro video using Veo 3. [00:07:34] S1: So they've doubled tokens since May and then yeah, totally [00:07:37] S0: The more interesting thing here is that Safe Super Intelligence, Ilia's small $32,000,000,000 [00:07:45] S1: Yeah. [00:07:45] S0: AI upstart is gonna exclusively use Google TPUs. That's exciting. That seems like a bombshell.
00:07:45 · Jordi Hays, John Coogan
[00:07:45] S0: AI upstart is gonna exclusively use Google TPUs. That's exciting. That seems like a bombshell. [00:07:53] S1: Maybe. I mean, you gotta get big infrastructure somewhere. And if everyone's optimizing against NVIDIA, maybe there's maybe there's like a slack capacity in TPUs. I mean, you're certainly like, if you are trying to bet on a hyperscaler to hit your horse too and you see that Google's increasing their CapEx pretty significantly to 85,000,000,000, like, you you know that they are going to have, like, enough capacity to surface to to to service you to the tune of one gigawatt or something like that if you wanna do some massive training run. I do wonder how how compute constrained SSI is in the sense. Like like, much is Ilia feeling the scaling wall? Because for a long time, it was scale is all you need. Just go bigger, bigger, bigger. And that's certainly the playbook that we've seen Sam Altman run with with the Stargate project and with the bringing in Masayoshi Son, bringing in Donald Trump and Crusoe, and putting together this, like, team that can actually marshal half a quadrillion dollars. Is that right? It's 500 no. No. Half a trillion. It's it's 500,000,000,000 They they wanna [00:08:59] S0: haven't marshaled it
00:08:59 · Jordi Hays, John Coogan
[00:08:59] S0: haven't marshaled it [00:08:59] S1: yet, John. It's a lot of money. Yeah. They're working on it. It's half a trillion. But but but SSI, at least at least like the the the rumors or the memes around it have been very much have been more focused on let's advance the research. And maybe that next phase of research is not purely compute bound. It's not it's not purely, hey, we're in this scaling race. We're gonna build a bigger transformer. That's that's honestly a little bit of what Elon's been He's saying, like, we have the transformer architecture. We're gonna scale this up a ton. And and where we are great is building a massive cluster, the Colossus, 100,000 [00:09:36] S0: Yeah.
00:09:37 · John Coogan
[00:09:37] S1: GPUs. We'll build it faster, and we'll get it up to speed, and we'll train it. And in three months, when most people would think it would take a year, we haven't been seeing that from SSI. S s s we haven't seen, oh, SSI has tents with GPUs in it. That that's their that's their edge. Yeah. The the the edge with SSI, as it's been pitched at least, has always been foundational research progress, just making making progress on on the underlying research that might be an elegant solution that increases the efficiency, and you get a better model with even less GPUs. But if they need a ton, Google will have them because they're investing more in CapEx more than ever before. While talking while taking a question on agentic capabilities, it sounds like Sundar said this, when we built our series of 2.5 Pro models, it's the direction where we are investing the most. That's definitely exciting progress, including in the models we haven't fully released yet. Long answer on agentic progress. The good news here is that we are making robust progress. We think we are at the frontier here. He said that they he said they have projects running internally, but now they are slow and expensive. They see the potential and are making progress on both. Gemini has a deep research product, which I would call an agentic product. Yeah. The question is, are they going to try and match what ChatGPT launched with agent mode sooner than later? Is there a security or a risk thing? Google's been a little bit more hesitant to just throw stuff out in the wild, but they're moving very quickly. And I think they will probably be able to get something up. There's certainly nothing on the CapEx or or model capability side that would stop them. And Elon Musk just says, wow.