01:04:24 · Jordi Hays, John Coogan, Stacy Rasgon
[01:04:24] S1: And I think going back to this quote from David Einhorn, he says, what we do know for certain is that the big tech companies that want the world to spend trillions have huge financial incentives to be believers. In case you haven't noticed, Wall Street is also being paid a lot to promote the story. And in that same letter, Einhorn and Greenlight said, this was this was the AI math makes no sense. [01:04:50] S0: Mhmm. [01:04:50] S1: It was basically the way it is today. Consumer business spends $1 on a ChatGPT subscription, is OpenAI revenue. Yeah. Then OpenAI provides the service by spending $2 on Microsoft AI infrastructure, which is Microsoft revenue. Then Microsoft spends $0.60 leasing GPUs from CoreWeave to handle the compute load, which is CoreWeave revenue. And then CoreWeave spends $2.4 on chips from NVIDIA and another $2.4 [01:05:18] S0: Yes. But you're completely discounting exactly how addicted the four point zero user is. They will pay any amount. So let's say that it's it costs $20,000 a day to serve a four o user, they will find the money. They will be stripping copper out of empty buildings to pay for their four o bill if they have to. They will be they will be Whatever [01:05:41] S1: it takes. [01:05:42] S0: They will be they will be breaking into cars to to sell stolen CD radios. [01:05:50] S1: Yeah. It is it is notable that Blue Owl has sold off 16% [01:05:55] S6: Mhmm. [01:05:56] S1: Past month. So even during this sort of like boom in lending Mhmm. It's not they're not getting very much credit for it.
01:05:56 · Jordi Hays, John Coogan, Tyler
[01:05:56] S1: Past month. So even during this sort of like boom in lending Mhmm. It's not they're not getting very much credit for it. [01:06:04] S0: Yeah. It's an odd it's an odd time. Like, there's still so there's still so many things that are working. The the results from that newcomer event, the AI is, Cerebral Valley, where he he had every it's so good. He he took, like, a straw poll on stage and asked everyone, like, who do you wanna short? And everyone said perplexity. So, like, he didn't really have to say, like, I'm shorting perplexity. He just was able to, like, take the temperature and sort of maintain, like, you know, some, arm's length distance to it while still, like [01:06:37] S2: Yeah.
01:06:37 · Tyler, John Coogan, Jordi Hays
[01:06:37] S2: Yeah. [01:06:38] S0: Getting the take out there. But everyone was saying that they wanted to buy more OpenAI, more Anthropic, more Andoril, a few other names came up, I believe. And and there's still, like, an incredible amount of bullishness in a lot of different areas. But, I think, yeah, people are still still worried about some of the some of the other stuff. But at at the same time, it feels like there's so many companies that have, like, gone through a they've gone through the AI pump and round tripped, not just Oracle, but even, like like, Klarna was going through the whole, hey. We're gonna get so many efficiencies out of AI, and then it came back to normal. But then they still got out, and it's, like, a reasonable company that's not, like it's not zeroed. Right? It's like, oh, yeah. Their business is just what it was before the AI boom, and they are not getting, like they didn't get moved up or down. And so there's just, a ton of companies that are like that. Where is Klarna since the IPO anyway? [01:07:37] S1: They're sitting, at around a 12,000,000,000 valuation, 12.6. Down Down. 20%? Yeah. Down 20 [01:07:46] S0: doesn't seem calamitous to me. That seems like pretty, like, solid. Like, I don't know, the overall market's quarter kind of, you know, up and down. I don't know. [01:07:57] S1: Affirm, on the other hand, is up 28%. [01:07:59] S0: How's Affirm doing? Up [01:08:00] S1: 28% Okay. In the [01:08:01] S0: past six months. In the past six months. [01:08:03] S1: They're getting Today
01:08:03 · Jordi Hays, John Coogan
[01:08:03] S1: They're getting Today [01:08:04] S0: is a real bloodbath. It's down it's down 6% today. Everything's down today. What else is down today? [01:08:10] S1: Bitcoin. Let's check. [01:08:12] S0: NASDAQ's down 1.3. Dow Jones, 1.3 as well. Well, we're we're going to ten thousand year mortgages. Doing ten thousand year mortgages. [01:08:23] S1: Announced the second round of stimulus checks already. [01:08:25] S0: Gemini three must save us. There there's a prediction on Polymarket that was quote posted by Sundar Pichai, CEO of Google. He says prediction markets are betting on Gemini three release week. 69% says November 22. Can't wait for that. We were wondering if we were gonna get it before Christmas. It was our Christmas present. It was Tyler's Christmas present. And, fortunately, looks like we're going to. And I couldn't be more excited that we are partnered with Gemini and Google AI Studio, create an AI powered app faster than ever. Gemini understands the capabilities you need and automatically wires up the right models and APIs for you. You can get started at a i.studio slash build. [01:09:09] S1: So Speaking of Google, capital says his final investing decision was to buy Google.
01:09:09 · Jordi Hays, John Coogan
[01:09:09] S1: So Speaking of Google, capital says his final investing decision was to buy Google. [01:09:15] S0: It's amazing. I think that's beautiful. He doesn't even need to say who he's talking about. It's, so obvious. It's Warren Buffett. It is it is beautiful that Warren Buffett is going long. Google, and, Darren here quotes this rune post that says, not enough people are emotionally prepared for if it's not a bubble. It's a good post. Like, it's like it's like, is it is it a is it a bubble if if all the big tech companies rip and there's, like, a couple neo clouds that trade down a little bit? Like, there's, like, you know, one or two application layer companies that burn a bunch of VC dollars, but there's still a new hyperscaler that's born. Like, kind of. I guess it's a bubble, but it's a survivable bubble. You know, it's just like, it's move on. You know? [01:10:02] S1: Yeah. Pulling bubbles.
01:10:02 · Jordi Hays, John Coogan
[01:10:02] S1: Yeah. Pulling bubbles. [01:10:04] S0: But people are getting wild on the timeline about Gemini three. Rune says the model must be good because the Google people have adapted the OpenAI culture of vague posting and hushed rumors and sending really weird texts. So we are ex and signals is LMAO explicitly calling out OpenAI culture of vague posting is hilarious. People are excited. There was also a a Reuters profile of Demis, and Morgan says, a Demis profile can only mean one thing. I imagine that that's Gemini three point o and that it will be good. The question is, like, what does good mean right now? Like, are we expecting anything that's, like, a a qualitative step function? Because my what what I would say what is Gemini three good? First off, I mostly just want better UI and little features in the app. Like, I want as a consumer product, I I just want, like, you know, better productization of the model that I already think is good. On the actual AI model side, I would imagine it's, you know, little 10% bump to how long it can reason. Maybe a model picker or, not a model picker, an automatic reasoning mode so that if I if I even if I think it's gonna take ten minutes to get me the answer, if it has the answer handy, it gives it to me in one minute, vice versa. It knows when to think really hard. It knows when to, think just for a little bit. I'm expecting it to be like slightly better at all the benchmarks. I don't know I don't even know what would blow me away.