Elon's Trillion Dollar Pay Package, Breaking Down the State of AI | Katherine Boyle, Mikey Shulman, Immad Akhund, Jordan Castro

28:03What else is in the timeline? There's so much in the timeline. Should we do our mega cycle review? Let's do it. I got the laser pointer. We got the laser pointer. our good friend Tyler Cosgrove has put together a slide deck for us that tries to help map the Mag 7. Really, I call it the TBPN top 10. The top 10. Well, technically nine. There's nine? Well, no, no. No, there are 10. There's 10. There's 10. We'll see. Well, one of them maybe shouldn't be there. No, no, no, no. No, there's actually 10 because it's the Mag 7 plus three more. Oracle. I forgot Oracle. Yes, you did. Okay. So it is the TBPN top 10.

28:43the 10 most important companies in AI, loosely, the Mag7 plus a few bonus ones. And we're going to try and take it through. And look, you're about to hit the horse. You've got a hoof, man. And so we're going to try and go through the various companies and rank them based on how AGI-pilled they are and how much they need AGI. Is that right? Yeah, basically. So let's pull up the actual scales. There's two axes. Let's go to the next slide. So basically on the horizontal, we have how AGI-pilled they are. Sure. So I feel like that's fairly self-explanatory. You kind of believe that AI will become something like it can produce the median economic output of a person.

29:23Yes, yes. And there's a few different ways to understand if somebody believes in AGI. It could be the rhetoric of the CEO or the founder. Sure. It could be the actions of the company. Yeah, yeah. Actions speak louder than words. Actions.

29:42And is there anything else that could lead someone to show that they believe in AGI? I guess it's mostly just the actions and the words, right? I mean, those are the main things that you can kind of do as a person. Okay. Say or do things. We will be judging them by both their actions and their words. So then on the vertical axis, we have how much they need AGI. Okay. So I think this is maybe a little harder. So I want to qualify this. Yeah. So, I mean, this doesn't necessarily believe that you have this kind of sentient AI that's as good as a person. But I think it more so in this context just means that AI will continue to become more and more economically valuable.

30:18Yes. To where you can kind of sustain building more and more data centers. You can do more and more CapEx. Yeah, there's a little bit of like, how much will this company be transformed if the AI wave plays out well? AI doesn't play out well. How chopped are they? How cooked are they? Exactly. That's a great framework. Yes. And then also, yeah, if we flash forward, nothing really changes. Total plateau, total decline in token generation or something. Is the business just continuing business as usual? Exactly. Okay. So who are we starting with? So let's start with Sam Altman. Okay. Sam Altman. Where is he on this?

30:56So Sam Altman, I think this is a pretty reasonable spot. Sure. He believes in AGI, right? He runs kind of the biggest AI company. He also needs AGI. Because if you imagine that, you know, if the models stagnate, right, they have a lot of capex they need to fulfill. If models stagnate, like, what are they going to do? Maybe they can, maybe the margins somehow work out, but you're probably not in a good spot if models get worse or if people start using AI less, if you're a same old one. But he's also not, you know, in the top, top corner. Okay. Right? And I think this is, you can justify this through a lot of OpenAI's actions.

31:30You see stuff like Sora. You see maybe the erotica. This is not very... Who's running the laser pointer? I'm running the laser pointer. You want them here? I maybe put them down here. Okay, explain that. Explain that. I think that more and more, at least in the short term, OpenAI looks like a hyperscaler. They're kind of a junior hyperscaler. And I think their actions are more... I think that OpenAI, a lot of people want to say that they're bearish on OpenAI at the current level. But ultimately, when you look at how their business is evolving, they seem to me like they'd be fine if the models plateaued.

32:11Yes, but I feel like the mood on the timeline was much more slide Sam to the left, doesn't believe in AGI. there was that post about like, if open AI really believed in AGI, they wouldn't be doing Sora, or they wouldn't be doing the erotica thing. All of those were very much like, needs it, but kind of accepts that it's not coming, and so stop believing it. That was the risk. I think in response to Jordi, I think, because it's not like this, the vertical axis is also just about like, if it's going to continue to be economically useful. So if people just stop using AI in general, or if people, if the revenue stops accelerating or any of this stuff, I think OpenAI will be in a bad spot regardless of the models actually getting much better.

32:58If they just make the models much more efficient to run, you could say that's not very AGI-pilling because the models aren't getting a lot better. But that's still... But I'm just saying if there was no more progress at all, we never got a new model from any of the labs, I think that OpenAI would add ads they would add commerce they would increase yeah so they might be fine without agi they would make agents a lot better but but i mean the 1.4 trillion enterprise but the 1.4 trillion in commitments like that is hard to justify if it's just the business today just growing like it's growing because it's just like it's good no no crazy breakthroughs like just like you yeah but they're, they're, they're going to play it out.

33:43The bull case, they're playing, they're playing crazy. You're laying out the bull case saying, Oh, if they just add ads, they're going to be able to hit the 1.4 trillion. No problem. That's not saying, I'm not saying they, they can pull back on a lot of these commitments. Sure. Sure. Sure. They don't like, okay. I don't, I don't think these are like going to end up being like real liabilities. Yeah. The business is just cooked. Okay. They can't hit them. Yeah. Right. So I'm just saying like, I think there's a shot they have. They could, you know, when they talk about having success in consumer electronics, which is something you talked about yesterday, they don't need – I think they can probably build a really cool device.

34:24Maybe it could be competitive with – if it can be at all competitive with an iPhone, they could be fine without AGI. They're also getting into – Let's get some other people on the board so we can see where people – Yeah, I think it's useful to be relative. because these are not quantitative numbers. So next we have Dario. Okay. So Dario is up here. So Dario is kind of, when you listen to what Dario is saying, he is, you know, he's extremely AGI pilled. Yes. Right. This is kind of the reasoning why he's so anti-China. Yeah. Right. Because he sees it as an actual race. This is going to super intelligence.

35:01It is nuclear weapons. It is a national, you know, security. Totally. It's a problem if China gets there first. Wait, what is that new sound cue? I don't think Tyler has sound effects. What is it? UAV online. UAV online. Okay, I like that. This is the UAV. We should give this UAV aesthetics for sure. This is good. Okay, continue. But there's also a sense that he needs AGI because if AI stops growing as fast and you imagine that things kind of settle where they are now, OpenAI is definitely in the lead. So you need a lot of continued growth for Anthropik to keep kind of making sense economically, I think.

35:37Okay, yeah, yeah, sure. Who else is on here? So I think next is Larry. Larry Ellis. Larry. Larry's in kind of an interesting spot here. So this is kind of a weird place to be where you don't believe in AGI, but you need it. Okay. How did you wind up there? So I think. You're probably wondering how I ended up in this corner. Record scratch, freeze room. So I think this is how you factor in. There's kind of the personal rhetoric, and then there's the actions of his company. Sure. So when you listen to Larry speak, he doesn't seem the type that believes in some kind of super intelligent god that is going to come, that's going to birth this new thing, and humanity will rise.

36:15But then you look at Oracle. Has anyone found his less wrong username? Is he on there regularly? I don't think Larry is reading Guern. Okay, got it. But you look at Oracle, and they need AI to work. They are levering up a little too much, or maybe not enough, depending on how AGI-pilled you are, but he's not very AGI-pilled. Exactly. It's kind of hard to square. It's a bold bet. Yeah, this is a unique spot, I think. It is. This is a unique spot. He's off the grid. Okay. Yeah. Who's next? So let's see. Who's next? Who did I? Okay, Satya. Satya, okay. There we go. I think this is a fairly regional spot.

36:55Yeah. Obviously, there's some sense where he is slightly AGI-pilled, or maybe more than slightly. believes in the power of the technology. Yeah, I mean, he's very early on open AI. He thinks that AI in general will become very useful, but maybe it won't become super intelligent. Maybe it's not going to replace every person. It's just a useful tool. It's the next... The quote I always go back to is him saying, like, my definition of AGI is just greater economic growth. So show me the economic numbers and that will be... It's a very practical definition. Yeah, I think people see him as very reasonable.

37:28He's not getting over his skis. Yeah, I like him in the center. I like him somewhere. That seems like a real... Yeah, he's also, you know, if AI doesn't work out, I think Microsoft is in a very good spot. It's going to stick around. You know, they're not crazy over-investing in OpenAI. They have a nice share of it. They'll write some code. If OpenAI works out, he'll do very well. If they don't work out, I think he's also... He's hedged. He's doing quite well. He's hedged. He's happy to be a leaser. Yeah, that's a good quote, too. You wouldn't be leasing if you were super AGI-pilled, right? You're hoovering up everything.

38:00You're keeping it all for yourself. Exactly. I think let's actually talk about that with, I think it's Jensen next. Okay. Yes. Jensen. So, so Jensen. He doesn't believe in AGI. He's the whole reason why we're here. Explain that. So yeah, this is maybe my personal take. Please. If Jensen was very AGI pilled. Yes. I mean, he is the kind of, he's the rock on which this all is built on. Yes. He has the chips. Yes. He has the chips. If he was AGI pilled, he would not be giving out those chips. He would keep them all to himself and he would be training his own model. Okay. So that's why I think he's, he's more on the doesn't believe in AGI.

38:28Yes. Yes. but if there's any kind of downturn in the AI economy. Could you put Sam potentially closer in this direction too because he's talking about getting into the compute reselling. Reselling, for sure. If there's so much demand and the models are progressing so quickly, wouldn't you want to just hold on to all that compute yourself? I think this would also be interesting to see over time. Sam has definitely shifted leftwards over time. He's moving. Basically this summer you've seen a lot of the actions of OpenAI They seem less and less AGI built Pretty much everyone has been moving the AGI timelines outward Which you could transform into no longer believing in AGI It's more just like the timelines have gotten longer this year Broadly, pretty much everyone There was a new blog post yesterday It was basically AI2037 There was a new one, it was AI2032 Wait, is it the same team?

39:24A different team, but the team of AI 2027 was promoting it. Yeah. AI 2027 should be like, it was actually AI 3027. We're off by a thousand years. Oh, typo. We're AI 3027. We couldn't get that domain. We were just off by a thousand years. Continue. Yeah, but there's definitely the sense where if AI, there's a downturn, Jensen, the stock is going to go down. Yeah, yeah. But it's not going to go down as far as Larry, right? because they're still not on it. They don't have insane capex. This is not financial advice. Yes. They don't have completely unreasonable capex. They're not levered up that I know of.

40:02Oh, yeah. Their Z-score is through the roof. Their Ullman Z-score is through the roof. They're looking pretty safe. Yes. Okay, let's see the next one. I believe, okay, Sundar. Sundar. He believes in AGI more than Satya, you think? Yeah, well, I think you can see this in kind of, They were even earlier in some sense than Satya, right, with DeepMind. So I think AI has played a fairly big part in the Google story very recently. They've always, I think, basically for the past 10 years, they've been trying to get into AI. Maybe their actions didn't actually do much, but, you know, they were the transformer paper.

40:37They, you know, Gemini now is... Applied AI. Like, they're actually applying AI for scientific discovery. They built self-driving cars. They built self-driving cars. And core Google search is just an AI product. It's just an index on information. They're organizing the... Also, compared to Satya, Gemini is a frontier model. Totally. Gemini 3 is supposed to be this incredible model. Everyone's very excited. Microsoft is not training their own base model yet. Yeah, and also there is a little bit of, if you really believe in AGI, the actions that we see are you squirming and being like, I've got to get in.

41:09It doesn't matter if I'm 1 % behind or 10 % behind or 80 % behind. I've got to get in. and I think we know someone who's doing just that. Gabe in the chat says, is this AGI or AG1 from Athletic Greens? Needs AG1. Needs AG1. Fine without AG1. Believes in AG1. Now, we're the only podcast that is not partnered with AG1. But we do like green. Green is our hero color. Yeah, that's true. Let's go to the next person. Okay, but also before then, I think Sundar is also definitely below this line because Google has been doing very well. AI was, at first, people thought of AI as like, oh, this is going to destroy Google.

41:50This is bad for Google. So if AI doesn't work, then Google is just in the spot they were before, which is doing very well. If AI does work, then Gemini is one of the best models. They'll do very well, too. Speaking of Gemini, Google AI Studio. Create an AI-powered app faster than ever. Gemini understands the capabilities you need and automatically wires them up for you. So get started at ai.studio.com. Yeah, continue. Okay, so I think, yeah, next is Zuck. Yes. So Zuck is also kind of in an interesting spot. Yes. I think Zuck is actually someone who has shifted rightward. It's fascinating. Yeah.

42:26I agree with this. You've seen this. He's been traveling. He's been traveling. Yes. Yeah, so for a while he was doing open source, which in some sense it's very AGI-pilling because you're building it, you're training a model, right? It's like you're moving the frontier forward. But it's also, it's open source, which you can kind of think as being, you're trying to commoditize everything. You don't think that there's going to be some super intelligence that will take all the value that you need to hold on to. You can kind of give it out. And now he's kind of moved towards closed source. We're going to get the best people.

42:55We're going to train the best model. It felt like Zuck was sort of like, oh, yeah, AI. Like, it's this cool thing. I'm going to check the box. I got my team. We did this fun little side project. It's this open source model. We kind of found our own little lane, but we're not competing in the big cosmic battle between open AI, anthropic, deep mind. Do you think that was just a counterposition way to try to win the AI war? Go say, hey, we're just going to try to commodify this market and take a Chinese approach? Yeah, commoditizing your companies is a good strategy. Yeah, that makes sense. And then maybe you could say that, oh, once the Chinese have been getting better, he can kind of step out of that position.

43:37Sure. And he's moved towards close source. Yeah. But now it feels like he's way, way going way harder on the AGI vision, paying up for it, investing so much money in it. Yeah. You know, it's like. And depending on how much he invests, you could see him neat pushing up. It's true. It's true. But right now, the business is just insane. It's so phenomenal that even if he winds up spending all this money and they don't even get a frontier model and all the researchers are like, yeah, we didn't discover anything and we're just gone. The business is fine because it's such a behemoth. So that's why he's fine without CGI.

44:15After earnings, they took a fairly big hit. So maybe he should be a little bit higher. Yeah, maybe a little bit. But I think MetaBroadly is still a very safe play if AI doesn't work out. And it was the same thing during the metaverse. It was like he believed in the metaverse. He invested in the metaverse, but he never needed the metaverse. And so after the stock sold off like crazy, it went right back up because everyone realized, wait a minute. But he's also raising debt off the balance sheet, which would kind of could push him up into this zone, right? If he's like, you know, if you're worried about, why don't you just carry it yourself?

44:46Why don't you carry it yourself? You believe in AGI, just let it ride. Yeah. Okay. Who else are we missing? Okay. So I think next is... is Jassy. Oh, Elon, Elon. Elon. Okay, Elon. So Elon, I mean, he's been AGI-pilled, I think, for a very long time. Super AGI-pilled. I agree with that. OpenAI co-founder. Even before that, I think he was fairly big in the safety space. Totally. You see him, even on Joe Rogan, he was talking about AI safety. He still believes it. He hasn't backed off. And AI safety is important because it's going to become super intelligent. It's going to take over the world. Totally.

45:18Yeah, even this was part of the dialogue around the new comp package wanting the voting power to be able to be part of securing his robot army. Oh, yeah. Interesting. Yeah, I think robots are very underrepresented on this board. Totally. And Elon is kind of the main player in that space right now. Yeah, so he talked yesterday about humanoids being sort of an infinite money glitch. And I feel like you kind of need AGI in order to kind of unlock the infinite money glitch. Yes, but at the same time, very strong core business. The cars don't need AGI. The rockets don't need AGI. Starlink doesn't need AGI.

46:01So he's not entirely indexed on it in the way the foundation labs are, right? Yeah. I mean, there's definitely a significant part of the Tesla market cap that is basically betting on robots. Totally. But there's also a big part of it that's just based on car sales. But that's why he's in the middle of the needs are fine without. It's like he is betting on it. Yeah, we say like XAI. He's not betting everything on it. He's not betting everything on it. Yeah. That's just one piece of the Elon Musk. AGI more than SpaceX. SpaceX seems very uncorrelated with the AI. Unless you have the data sets. He's happy to be a telecom baron.

46:34He's happy to be the new Verizon. Who else is on here? I believe now is Andy Jassy. Jassy. He is, I think broadly does not believe in AGI, although he fairly major stake in Anthropic. Which maybe is very AGI playing, but they have not. He's hedging with Anthropic. Yeah, I think that's basically what you can say. He doesn't seem at all worried about kind of the core business. AWS seems to be... And isn't Google building their position in Anthropic? I saw some headline about that. That was wild. Imagine. Sundar, such a beast. Sundar and Satya, really, both of them. They just have huge stakes. Broadly, Andy Jesse seems very quantitative.

47:15He's focused on the numbers. Realistic. He's not making these grandiose statements about the future of... of intelligence or what humanity is going to be like. Yeah, when you look at the AWS earnings, it feels like they invest when the demand shows up and they build data centers when there is demand and they are not in the business of doing, you know, a 10-year hypothetical sci-fi forecast based on breakthrough. So Ben Thompson on the recent Sharp Tech, we were listening to it on the way in, he was talking about how Amazon has repeatedly said, hey, we're supply constrained. We have a lot more demand than we can fulfill.

47:52With this new deal with OpenAI, the$38 billion, which only got announced Monday, which, again, feels like forever ago at this point. But Ben was kind of hypothesizing that they kind of let OpenAI jump the line because that$38 billion deal was starting effectively immediately versus some of the other, like Larry's deal with OpenAI was announced. It's this massive backlog is revenue that cannot be generated in a meaningful way today because they have to build the data centers that can actually deliver the compute. Yeah, I mean, and Amazon has been building data centers. They're building data centers.

48:32They're basically, they built all the data centers for Anthropoc. But it still seems very kind of restrained. It's not overly ambitious. Anthropoc has had issues with capacity. And it's probably because Andy Jassy doesn't want to get over his skis. He doesn't want to build too much. So I think that's why he's kind of on the left side. And also, I mean, if you think AI doesn't work and we're going to kind of be in this, you know, the same spot of Web 2.0, you need AWS. You need your EC2 server. I think he's very well positioned there. And then I think the last one is Tim Cook. Tim Cook. Whoa. Let's go.

49:07Let's hear for Tim Cook. Criminally underpaid, but has done a fantastic job. not getting over his skis. Yeah, this one I think is fairly self-explanatory. I mean, he seems... He doesn't believe in AGI. He doesn't believe in LOMs. He doesn't believe in chatbots, apparently. I don't know. The new Gemini deal signaling. As of two years ago, Apple was like, yeah, we're not doing that stuff. Yeah, there was that famous... Three years ago. But I think what's under-discussed, my takeaway from our conversation with Mark Gurman and Apple's ambitions around their new LLM experience with Gemini is that I think that there's a very real scenario where Apple wants to compete for the same user base as OpenAI.

49:57They want that transaction-based revenue, that commerce revenue. Yeah, I disagree on this point. I'm not saying that I'm putting their odds at winning that market at over like 10%, but I think that they would be right to realize that it's potentially an opportunity. Yeah. Is it a billion dollars a year that they're paying Google for? I believe that's the number. Yeah. That feels really low, doesn't it? I don't know. It just feels super low to me among all the different deals that are going on. When I think about like, when I just think about like the value of AI on the iPhone and you're like$1 billion.

50:41Like when I, when I think about like what is the value that we're, that the market broadly is putting on like AI for the enterprise, whatever. When Sundar on the Google earnings call was talking about their top, I think 10 customers and how many like trillions of tokens they were using, but it really was netting out to like$150 million of like actual revenue. And so this is, this will be their biggest customer for Gemini on day one. Until one of our listeners gets on. Of course. Yeah. And I think, I mean, even with the Gemini news, Apple still seems very kind of reactive to AI. Yeah. They're not kind of seeing, oh, this future where AI is going to do everything and moving there right now.

51:24Yeah. They're kind of seeing where the demand is, seeing where the users are, and then moving, which I think is very kind of classically, that's how business works. It's not very AGI-pilling. Also, on that point of$1 billion, I have no idea. How can they possibly know the amount of inference that's going to happen in Siri plus Gemini over a year period? There's just no way to predict that. Or can they? I feel like if the integration's good, there will be a ton of queries. I didn't read the$1 billion headline. And Chad can correct me if I'm wrong, but I didn't read the$1 billion headline. I felt like that was a technology license, not necessarily inferencing.

52:07And then there might be consumption on top of that. Yeah. Okay. Or maybe, yeah, yeah, yeah. Maybe they're licensing Gemini and then they pay per query for the energy and the CapEx. And couldn't a lot of the stuff be done on device? For sure, yeah. Because Gemini, Gemma has like baked down models that could be smaller. So that's possible. Chad says, where's Tyler on the chart? feels like Tyler isn't AGI pilled anymore I actually am on the chart here let's go to Tyler I am I think I'm very AGI pilled I'm ready for the Dyson sphere I think it's only a matter of years handful it's a couple thousand days away you need to publish a blog post AGI 2026 AGI 2025 we still have a month left AGI on Christmas This Christmas.

52:58It's coming this Christmas. This holiday season. But why do you need AGI? I think I also need AGI. Why do you need AGI? Well, I mean, look, if you look at the current jobs data for college-age students or post-grad, it's looking pretty bad. It's bad. So I think you kind of need AGI to really boost the economy. If AI does not work, the macro economy is looking not good. Oh, sure, sure, sure. So I feel pretty bad about my job outlook without AGI. Sure, sure, sure. Even though you're already employed. That's hilarious. Well, thank you for taking us through. Chat wanted to know, where would you put Lisa Su?

53:37AMD. Yeah, so Lisa Su, I think she's probably been quite reactive, actually. You've really only seen AMD start kind of making... You don't get credit for being reactive, is what you're saying. Yeah. I think it's really only over the past year, maybe, that she's been making any kind of deals, right? You saw George Hots maybe a year or two ago basically trashing AMD chips for how bad they were. But maybe she's given so much of the company away, maybe she thinks that shares won't have very much value in the future. She's like happy to just give away 10 % of the company. Yeah, so I think I would, if I had to pick somewhere, I would say that she is, she's honestly getting a little close to Larry.

54:21in that AMD, it's very much like an AI play still right there. Chip company, obviously. But you don't get the feeling that she's a true believer. Yeah. That tracks. Okay. That tracks. The new Siri, I'm reading through some of Mark Gurman's reporting. Speaking of Apple and AI, the new Siri is planned for March, give or take, as has been the case for months. Apple has been saying for nine months, it's coming in 2026. Apple simply reiterated on that. And then on the actual deal, it's a$1 billion deal for a$1.2 trillion parameter artificial intelligence model developed by Google to help power and overhaul the Siri AI voice assistant.

55:01The two companies are finalizing an agreement that would see Apple pay roughly$1 billion annually for access to Google's technology with the Google model handling Siri's summarizer and planner functions. Apple intends to use the Google model as an interim solution until its own models are powerful enough and is working on a one trillion parameter cloud-based model that it hopes to have ready for consumer applications as soon as next year. That is interesting. I feel like they're going to have to pay a lot more to actually run all the queries, generate all the tokens. But I'm sure we'll find out more as Google reports earnings and Apple reports earnings.

55:42Before we get to that, let me tell you about graphite.dev. Code review for the age of AI. Graphite helps teams on GitHub ship higher quality software faster. What else is in the news today? We have some breaking news. What's the breaking news? TJ Parker has finally found a new car that he likes. Huge. Guess what it is. Guess. A new car that he likes. It's not the Cerato. It's not the Lamborghini Huracan Cerato. No. It is the Ford Raptor R. Look at this. Look at this. Finally found a new vehicle I quite like and great gas mileage to boot. This was the problem with your Ford Raptor. It wasn't the R.

56:23It wasn't the R. You only had 600 horsepower instead of eight. That was the main problem. I also needed to park it in cities, which is just absolutely brutal. Yeah. What are the specs on the Raptor R again? It's pretty wild. How much horsepower? Raptor. Zero to 60 and 3.9. 720 horsepower. 5.2 liter supercharged V8. Fantastic. It's kind of the Julius.ai of trucks. It really is. The AI data analyst. Connect your data. Ask questions in plain English. Get insights in seconds. No coding required. I'm trying to go through the timeline. We got Wilmanitis talking about some other stuff. Let's pull up this post from Pegasus Since it's relevant to The overview we just did Pegasus says Altman today Of course he's talking about yesterday We're looking at selling compute but we need as much as possible Zach last week we could sell compute Are we in a compute Shortage or not Because both are saying they're buying as much of it as they can And thinking about selling it That's very interesting Yeah I mean the supply and demand dynamics I like that story about Jensen where he understands the dynamic here, but there's still, you get kind of crushed if there's a sell-off in terms of demand.

57:46You just have to go back and study what was going on during the crypto boom for NVIDIA. During the crypto boom for NVIDIA, it was like, buy as many as possible. Just ramp, ramp, ramp. And then they literally had a glut. Well, yeah. Completely had a glut, and they took huge write-offs. Yeah, the steel man here, if you're a large tech company and you have the ability to and the financing capital to buy a lot of GPUs, it's not the worst idea to buy as much as you can so that you have preferred access to it and then resell some if you have more than you need, right? Yeah. But yeah, the question is - Historically, that is not how things have been done.

58:29Like, I understand the pitch there, but, I mean, we just talked to David Buzuki from Roblox, and he was saying that, look, we had our own on-prem, but then we had spikes of demand. And so we went to the hyperscalers for that because they can load balance across, well, people are playing Roblox here, and then maybe they watch some Netflix over there, and they are storing, you know, all sorts of different data. And there's different workloads that happen at different times. And so traditionally, the hyperscalers have been able to service multiple users across them. Jumping straight to selling compute, I think that the timelines are a little bit funky on this one.

59:17It seems odd. It seems rushed. It seems rushed, especially when your core product is growing so quickly. Like Google Cloud Platform was released, what, seven years after Google launched? Same thing with AWS, same thing with Azure. Like these were very mature businesses, very stable businesses with cash flow that were able to justify the investment. It was not something that was done as part in the growth phase as much. I don't know. Speaking of other wild, wild, you know, new business lines, Elon apparently confirmed that Tesla is going to build a semiconductor fab. Yep. I've been advocating for this for a long time.

59:56I'd love to hear what Elon had to say. But first, let me tell you about Fall, the generative media platform for developers. The world's best generative image, video, and audio models all in one place. Develop and fine-tune models with serverless GPUs and on-demand clusters. Let's play Elon. What do we got here? Let's play Elon. production from our suppliers, it's still not enough. So I think we may have to do a Tesla TeraFab. Whoa. Great name. That is a bombshell. It's like giga, but way bigger. Tera, yeah. He's feeling good right now. I can't see any other way to get to the volume of chips that we're looking for.

1:00:44so I think we're probably going to have to build a gigantic chip app. Morris Chang at TSNC is like, no, actually I'm fine. I will supply you. Samsung is like, I'm good. I will do it. I'm good for it. Pull up this other video of Optimus. Even when it was crazy. Because there was some new moves of Optimus getting jiggy. Let's see, it is at the bottom of the Tyler app. Look at this. look at this okay so this was shown in contrast to the first Optimus which was just a guy in a suit this thing has motion yeah it's not bad this feels like getting pretty close to unitary level definitely at that level I don't know imagine you're working late at the office one night and this thing just walks out onto the floor and starts looking at you and doing these moves.

1:01:43Yeah. Weird. It's so, it's so weird how these new projects, like they seem, I understand why a lot of people look at this with skepticism, but at the same time, like it just doesn't seem that complicated to just manufacture that thing and ship it. Like, like they do it at, you know, in China, they do it all over the place. This doesn't seem that insane and at the same time like apple couldn't ship a car so like these new projects like they do kind of elon's shipped some cars though yeah yeah uh i i meant it more as like as like if you're doing one thing it can sometimes be hard to branch out into the other thing i remember a few weeks ago there was a headline around uh tesla ordering like 600 million dollars worth of actuators yeah like it seems like they're going into some sort of production yeah um i wonder what who they're gonna you don't need 600 million for like test just for testing it's just a the the car analogy is so tricky because it's like uh most people that bought teslas already had cars right and so it's just like a one for one slot swap who are you replacing with this?

1:02:59You know, it's tricky. What do you think? Well, with this, you can replace, uh, like an exotic dancer. Yes, that's true. I don't know if many people have that, an exotic dancer in their life who they're ready to, who they're ready to replace. Um, it seems, I don't know. It'll just be very interesting to see where these things actually diffuse. Like, because if you sell them, if you sell them for one price to consumers to do their laundry, but they work in an industrial capacity, like people will just buy them and use them for that. But when we talk to people who are in industrial environments, they're like, I definitely don't want a humanoid robot for that.

1:03:39I'd rather just like a big, massive robotic arm that's like bolted to the ground and can actually lift like 10 ,000 pounds as opposed to one that can only lift like 50 pounds. I don't know. So Ryan says they're going to sell them to the police. We'll see. Maybe. They certainly would be, I don't know, backup. I think they would be pretty effective deterrence. Maybe. I like Vittorio having some fun on the timeline. Going viral, getting community noted. Vittorio said, Elon Musk now has$1 trillion in his bank account. That's 1 ,000 times$1 billion. He could give every single human on Earth$1 billion and still be left with$992 billion.

1:04:19Let that sink in. People love this funny math whenever it drops. What else is going on? The humanoid. Yes, China has a similar humanoid robotic project, although it's way, way scarier because it went full Terminator mode on this. Of course, this one is hooked up to power. But... Yes, so this video was in response to... Earlier this week, there was a... I think it was a unitary presentation and they brought out this new robot and it was walking with such like swagger, natural gait. Yeah, basically that people thought it was actually just a person in the suit. And it was not. And so they replied with the, fully stripped down.

1:05:00That is remarkable. Yeah. Oh, because they put it, they put it, they put like the suit on the robot. Yeah, it was like a Neo. Ooh. Kind of like the 1X robot. We're getting spooky, spooky territory. This is pretty crazy. Oh, wow. Oh, yeah. It does look human. It does look human. I'm looking at the other video. Sorry. Let me tell you about Turbo Puffer. Search every byte. Service vector and full-text search. Builds from first principles and object stories. You see these new logos, John? Fast, NX, cheaper. You see these new logos? Oh, wow. Stacked. Stacked. Absolutely stacked. They're cooking. Chris Backey says, In the last 10 months, three very talented friends have joined separate, hot, early-stage startups in senior roles and quit after realizing that the company's actual revenue was significantly less than what the founder had told them during the interview process and shared online.

1:05:52Hmm. Few things to clarify. I'm not joking. In two of the three cases, it would be hard to tell that you were joining a bad actor company on the surface. Solid investors. Founders worked at unicorn companies. The third was maybe more obvious. The only ones who really lose here are the employees. The investors have 200 portcos and can afford some losses, and our industry doesn't really pursue founders over fraud slash misleading information up to a certain point. Four, if you're thinking about joining an early stage startup, ask to see their Stripe and or signed contracts before you join. Not bad advice.

1:06:26Yeah, again, this going back to spring, right? That just felt like there was this, like, every founder was feeling this insane pressure to show like one to 10 million of, like a one to 10 million ramp that was just insane. Yeah. And the weird dynamic is that like, as that pressure ramps up, you just get more and more incentive to fake it with a community adjusted ARR and contracts that don't actually stick and all sorts of different twists on something that's like, is it actually cash? Is it actually people coming in? Are they on long contracts? The quality of revenue has been maybe degrading, but also just maybe in just, you know, it's just been easier to game than ever.

1:07:16There's been more incentive to game it than ever. So stay safe out there, folks. Kazakhstan has signed an MOU to buy up to 2 billion of advanced chips from NVIDIA. Let's hit the gong for Kazakhstan. Warm it up, warm it up, warm it up. Boom, great hit. great hit for our friends in Kazakhstan good to see them getting into the game does Borat take place in Kazakhstan isn't that is that the whole thing maybe they should do Borat 3 about data centers he he goes to a data center Sasha Baron Cohen would be on the AI reprises his role all to bring AI to Kazakhstan Sasha Baron Cohen doing a doc like a like yeah like on just on AI incredible incredible.

1:08:04It'd be amazing. So I can use it to do better web search. So I can use it as an autocomplete. You're not doing the Borat voice, brother. I'm not going to do the Borat voice. I will do the profound voice. Get your brand mentioned in chat GPT. Reach millions of consumers who are using AI to discover new products and brands. Go get a demo. The Kobe SE letter says breaking NVIDIA's losses accelerated to negative 5 % of the day, now down 16 % since Monday's high. That marks a drop of$800 billion since Monday. Wow, that is a wild sell-off. Didn't Tyler quote this and say, is this bullish or something like that?

1:08:47No, that was on the DoorDash. They went down 20%. Everyone's down 10 or 20%. Yeah, whether you beat or miss, you're going down. Wait, is this true, Rigov? Ari Emanuel said he's working with Elon to bring Optimus to the UFC? That would be incredible. Please don't be fake news. Please don't be fake news. We need robots in the UFC. I mean, Rogan and Palmer Luckey were talking about it on their podcast, and it certainly seems like a hilarious and wild thing, even just as an exhibition match before the real UFC. Get an Optimus. though here's a question would elon want a human to thrash on optimus is that good for his brand or would he want optimus to win and if optimus wins then is that actually entertaining i don't know i think putting optimus up against the best mma fighters in the world yeah is fine it's impossible though right because if you just kick metal you'll just break your foot and if you so This was Ari Emanuel at the All-In Summit.

1:09:52Yeah. It was released, I guess, one day ago. Oh, interesting. But he was talking about wanting to break. Wait, wait. What a weird, what a crazy, crazy timing. That's so interesting. He says, I saw what he's creating. The man's a genius. I want to do a UFC fight with his robots. Yeah, yeah. Very cool. Heisenberg is sharing a little bit of red here. Microsoft down 10 % in the last eight days. Nvidia, 12 % in the last four days. Malantir at 16%, down 16%. Meta, down 18%. Losses have accelerated. Compound248 was sharing yesterday. He's like, when I clipped this, I didn't expect to crash the markets globally.

1:10:42but he certainly that clip seems to have played a part in this little sell off. Always take a victory lap for being for having world historic consequences I actually think that that particular clip was like less than 5 % of the total views on that clip because there were other people that clipped it and it got but always claim victory that's the lesson. That's right. Anyway we have our first guest of the show Catherine Boyle from Andreessen Horowitz joining us. She's in the Restream Waiting Room. Welcome to the TVGN Ultradome, Catherine. How are you doing? It's great to be here. Good to see you guys.

1:11:19Why are you playing the night vision goggles sound? You already got a new soundboard today. I felt a little tactical. I felt a little tactical. It's great to see you. Good to see you. Should the Department of War get a soundboard during these speeches? I heard Hegseth was giving a big speech. Are they sticking mostly to the cheers and the clapping? Or are we firing off muskets? Are we firing off cannons? Is there pageantry alive in the Department of War speech? Bring the pageantry back to the Department of War. That is important. You guys should definitely have your own sound for them. For sure.

1:11:55You guys are getting the first look at a major speech. Yes. So it is just ending right now. I just hopped off the live stream. And I can tell you my phone is blowing up with just how many people are enthusiastic, not only in the venture community, but just across Washington about what has been said in this speech. And it's going to have dramatic repercussions for defense tech and for Silicon Valley and the broader defense innovation base. So it's an extraordinary speech. You'll be hearing a lot more about it, but you're getting the first look. Fantastic. Take us through it. And it is definitely, I'd say, you know, for the last, say, 10 years or so, we've been talking about defense reform, why it is so important for increasing competition for startups.

1:12:38And what is amazing about this speech is that it's not just lip service to, yes, of course, we need competition. Basically, what you've been hearing in Washington for the last 10 years, it goes line by line. I mean, Seth actually made a very, very funny comment. He said, if you're watching this on Fox News, your eyes are glazing over because this is the most boring thing you've ever heard. But for everyone in the room, from the Andurls, the Palantirs, the SpaceX's of the world, all the way down to the little guys to the primes, it was like Christmas come early, right? Because it really is changing the way that acquisitions are going to happen inside the department in ways that are so meaningful, not only for Silicon Valley companies, but for private equity-backed companies and for the primes.

1:13:17He basically told the primes in the very beginning, you have to be investing more in research and development. The companies that have been getting the large programs cannot just continue as business used to be done. He actually quoted Donald Rumsfeld, didn't tell us he was quoting Donald Rumsfeld for the first five minutes, and basically went line by line through a speech that he gave on September 10, 2001, about how important it is to change the defense acquisition process. And he said, you probably don't recognize this, but the speech I'm giving now is actually Rumsfeld's. Wait, September 10th?

1:13:48Yes, yes. The day before just happened to be? Happened to be the day before where he said we have to change the application process. That is a crazy coincidence. Very crazy coincidence, but it shows you just how long this has been going on. Sure, sure, sure. So I'm happy. I have my notes here because I was taking Komi's notes throughout the speech. Yeah, let's go line by line. To let startups know what's going to be changing. Yes, please. But the first thing that I think is really important for all of the startups in Silicon Valley is that he mentioned and called out the importance of commercial-first technology.

1:14:22So this is going to shock people who are not in the defense world, that right now, if you need to build a product specifically for the Department of Defense, the vast majority of contracts that are given out are not given out with a purview of what exists already in the marketplace that we can buy. Yes.

1:15:07will buy it or that that is that is the goal and then we will turn it into what we need to for actually for the war this is cots versus gots for the dod dow were nerds like commercial off the shelf versus government off the shelf uh this is what andrel and all the defense tech community has been beating the drum on for basically a decade now and it feels like it's finally getting echoed back from the administration is that correct yeah yes and there was there was a lot The administration and the EO that was out a couple months ago said that they wanted commercial-first technology. Now you're hearing it straight from the secretary's mouth.

1:15:41There's always dust-ups about this on Capitol Hill with various parties arguing over whether we need more studies or whether we should just do this. It looks like now this is real change that's about to happen, and it's going to be transformative for all of the startup community. What it really says is that the best products will win. Yeah, who are the losers here? Well, I think, I mean, the major loser from this speech is the primes. I mean, he did. He said, I don't want to pick on anyone. I'm not going to call out companies, but you have to be investing in research and technology. Like you have to be the fact that these that the primes only spend two percent on research and development and are focused on, you know, constantly kind of maintaining the status quo.

1:16:22That is not going to happen on his watch. I think it was two percent on R &D. Yes. And haven't historically been been acquiring venture back then. company. So this is also the problem where they're really just on a much slower timeline than is needed by the department. And I'd say that the takeaway from the speech, and I think the sound you'll be hearing more and more about, is that the new process is really going to be used to enable speed and volume. And one of the things that I think a number of companies have been talking about for a long time is that by taking more risk in the acquisition process, you take less risk on the battlefield.

1:16:58And he actually echoed that. He said, we want to take more risk. We want to have, we want to reward the people inside of the department who are making risky decisions or supposedly risky decisions because they're acquiring new technologies or from companies that are not Raytheon and the kind of big guys, right? But what they ultimately said is like, they want those, the people who are going to be making those acquisition decisions to be taking on the risk because it takes less, it will affect the risk of the warfighter, right? It will make things safer for the warfighter. So it's basically linking this entire process, bringing it back to what needs to be done before war start and before products get into the hands of the warfighter.

1:17:37This is like a top-down directive. And then who are the players that need to actually put this into, like, who are the kind of key players that need to put this into practice? Yeah. So, I mean, like, the main thing that I, this is definitely a message sent to Congress, because there's currently, you know, two different bills going through on Congress that are actually pretty aligned in most of the things that they're saying. There's a few call outs where I think you'll see sort of some sort of back and forth between the House and the Senate. But this is definitely a call to them. Like this is happening on the secretary's watch.

1:18:08This is something that he cares deeply about. This is something that the administration cares about. And so it definitely is a huge push, not just a major signal, but they sort of outlined this is how we're going to change things inside the department. One of the biggest things that they do is they created this sort of portfolio-based approach of how they're going to be allocating capital to companies now. And this is always shocking when I explain to people that this is revolutionary because you'd say in a normal company, you're given a budget. And if the vendor you choose fails, you'll go to another vendor and you'll use that capital in order to buy whatever product you need.

1:18:42That is not how the department works. And so with this new portfolio-based approach, they're basically saying we're going to give what's known as the PAEs, the Portfolio Acquisition Executives, the right to say, okay, if this one product that we acquired 60 days ago isn't working on the battlefield, we can use the rest of the money and reprogram it for a product that is working. And right now, that is not possible inside the Department of Defense. Right now, you have to write a whole other list of requirements and ways that you would build a new product in theory that can take years to actually build.

1:19:13And by the time it's built, it isn't meeting the requirements or the needs of the warfighter. So the fact that they're even changing the budgeting process and that they have an explicit call out of how they're going to do that, that will show up in this year's NDAA once it's passed, and it will be the way that things are done from now on. So this is extraordinary. Yeah, so this means that if a contract is allocated to a certain company and they start underperforming in real time, the person who's in charge of allocating that budget can basically say, hey, this other startup, we're going to rotate this budget over here because they're actually going to deliver on this.

1:19:51I've heard some horror stories and even opportunities where some of the bigger primes have just held on to contracts forever. They're kind of like barely delivering. They're kind of like trying to deliver and startups coming in and saying like, hey, we'll actually take this over. And sometimes those deals can happen, but this seems like it'll be a lot more effective. Are there any new categories that are opening up generally? I mean, we've heard this week about data centers in space. So if you're tracking data centers or space, that's a story. But I remember Secretary of the Army was talking about modularity of weaponry, taking apart a drone, fixing it.

1:20:35I feel like the whole idea of drones or autonomous submarines, these are new categories that sort of opened up. There have been a number of competitors. But at the thematic, like within defense technology level, are there any categories that you think this new administration is maybe more excited about than ever before? Yeah. So, I mean, you mentioned modularity and what the Secretary of the Army is talking about. I mean, this was actually called out in Secretary Hegg's speech where he said part of the reason why we need these PAEs and what were formerly called PEOs, but PAEs, these portfolio-based approaches, is because right now, if you want to take something apart and put it back together, you have to build an entirely different system to acquire those parts.

1:21:18And you have to go through a requirements process. So if you're building modular-based approaches that update in real time, and he actually said, like, we want it to be more like software. We want to do software updates in real time and not have to go through a requirements-based approach. You need to be thinking about things in terms of modularity. So I think there's a lot of different, even just changing the way the budgeting works, It opens up a lot of different technologies that were not able to participate in the acquisition system because they didn't want to go through the requirements process or they knew they wouldn't meet the requirements that are being put out by the Department of War.

1:21:54So it is going to radically change, I think, not only the types of products you're seeing in the hands of the warfighter, but just how quickly they get there. Makes a lot of sense. Switching gears a little bit, Sam Altman suggested on Wednesday this idea of, like, my interpretation of it was kind of like a GoCo model for AI factories. Without commenting on that specifically, can you give us the history of GoCo's, when and how they work well, and what kind of some of the challenges have been? Well, I actually didn't see what, can you give me a little more context? Yeah, so I'll read exactly what he said.

1:22:37So this was in response to clarifying the sort of backstop gate. He said, so anyways, he basically said, obvious one, we do not have or want government guarantees for open AI data centers. And then like later in the post, he said, what we do think might make sense is governments building and owning their own AI infrastructure. And then the upside of that should flow to the government as well. We can imagine a world where governments decide to offtake a lot of computing power and get to decide how to use it. And it may make sense to provide lower costs of capital to do so. Building a strategic national reserve of computing power makes a lot of sense.

1:23:13But this should be for the government's benefit, not the benefit of private companies. And just in some of the ideas around this just reminded me of how GoCo's have worked in defense tech, where the government is basically saying, like, we're willing to commit resources and land to, like, a certain initiative, and then you can have the private market potentially pay a role. And I have no idea how this would look like in the AI context, but I just thought it'd be helpful to understand, like, how these sort of, like, public-private partnerships have worked in the past, specifically in defense. Yeah, well, no, I can talk to one that was actually discussed today.

1:23:52So the Secretary of the Army was on Squawk Box this morning talking about specifically if we are giving the land and security for massive data centers, could the Army have its own data centers and be one of the largest providers and actually offtake that compute? So I do think that there are ongoing conversations. Candidly, I haven't seen massive conversations about how it's going to be enacted by the Department of War. But I thought it was very interesting that the secretary of the army was basically saying that this is something that they would be very interested in, that they know that they are going to need their own compute.

1:24:23And that if there is sort of a tradeoff of we provide the land and the security, there should be an understanding that that compute then goes to the army. So I don't know that they fully fleshed out the details of those sorts of things. But I do think there's a lot of conversations about this. And, you know, it's good to hear that Sam is also talking about that in certain ways. But I do think for the Department of War, they know that it's top of mind. And I think they're being a lot more experimental, which, again, comes back to if you have a system of requirements that is far more sort of open, I'd say, to experimentation or working with new companies that can do that.

1:24:57It does give the secretaries of the Army or the Navy a lot more authority to make those decisions on kind of the actual things like compute or energy that they need. Yeah. How do you think about the flow between entrepreneurial ideas, like problems that are identified in the private sector? Somebody starts a company. With Anderle, it's like the flipped model, this idea that R &D should maybe live outside of the taxpayer dime, be funded by venture capitalists. That company gets built. eventually those ideas are percolating into DC and then the world updates, America updates, we change the way we work, things get better, but and real benefits and people are pointing the finger, you know, like, hey, you're the one that advocated for this.

1:25:46And I feel like that's kind of where Sam's caught right now, where OpenAI is advocating for things that I think are good for America, but also might benefit OpenAI. And so he has to do this delicate dance of like, well, I was independent on this one, but this one does benefit me, et cetera, et cetera. It's like such a communications challenge. Do you think it's like, do you have any best practices for like how to not get over your skis on that? Well, I think it's, I think, as you said, it's sort of a typical Washington story where in order for things to change, you do have to sort of, you know, kind of Washington's a meme town.

1:26:20Just as Silicon Valley is a meme town. You have to be able to tell your story, why it's so important. I would argue that Andrew was exceptional at really telling the story of why you need attributable systems, why you need modular systems. Basically, the kind of frameworks that we're using today inside the Department of War came from those memes being established. But yes, if you're successful inside of Washington, you're going to tick off a lot of different people. Totally. And you're going to certainly upset sort of the bureaucratic class. And what was very interesting, too, about this speech is that the secretary actually called out the bureaucratic class and said, like, this is not a problem with people that we have.

1:27:00This is not a problem with innovation or technology. It is the bureaucracy that is stopping us from making great decisions. And I think that is something that has always been the problem with D.C. And I think the most important thing that companies can always do is just continue saying what they believe, going to Washington, building those relationships. And I think, if anything, over the last 10 years, this defense acquisition reform, which everyone said was impossible, is now actually at our fingertips and going to happen because it makes sense. The argument has been made. It's been made in dozens of different ways across many administrations, and everyone's in agreement on it now.

1:27:37So for founders that feel like, oh, gosh, look at these great companies that are getting backlash because they've kind of spread their own gospel and now they're being attacked for it, that's actually, I think, a sign of progress. If you're being attacked for what you're advocating in Washington, that's usually a good thing. Yeah, no, that makes a lot of sense. Final question. What is your most up-to-date guidance for your portfolio companies around the shutdown? Yeah, no. So, I mean, it definitely is having an impact, right? Like the companies that were expecting contracts to come in or reprogramming dollars, like they're not going to necessarily see it on the timelines that they thought maybe a quarter ago.

1:28:16That said, I do think that a lot of these provisions that are happening right now are so transformative that in the long run, like the kind of if we look back on what happened in the year 2025, it will have been a very, very good year for startups. Maybe their numbers will lag by a quarter just because, OK, like the funds can't be delivered depending on how long the shutdown goes down. I think it's having a greater impact actually on the citizen, right? I mean, what's happening with air traffic control, what's happening across civic benefits, that is something that I think we're now all feeling the pain of, or people who are dependent on those benefits are certainly feeling the pain of, or people who are traveling or planning to travel for upcoming Thanksgiving are certainly going to feel the pain of the shutdown.

1:28:58but it's certainly having an impact on companies. And I think most companies are still working exceptionally hard on the things that they can work hard on now, even if the, even if the contracts can't get signed or the press releases can't be put out or the dollars aren't coming in yet. But I think when we look back this year, the story of this year will actually be the defense acquisition reform and what's happening for these companies in the long-term versus sort of the short-term. Yeah. Short-term pain, long-term gain. Yeah. Awesome. Well, thank you so much for taking notes and giving us the full breakdown.

1:29:30Thanks so much. Yes. Always great to catch you. Have a good one. We'll talk to you soon. Thanks so much. Have a good one. Let me tell you about Linear. Linear is a purpose-built tool for planning and building products, meet the system for modern software development, streamline issues, projects, and product roadmaps, and start building. And we have our next guest, Mikey Shulman from Suno in the Restream Waiting Room. Here he is. Welcome to the TBP and Ultradone. How are you doing? It's great to be here. I'm doing great. How are you guys? Fantastic. Fantastic to have you. I'm very excited about this.

1:29:58I have a million questions, and I know that we're just going to go super deep into all the details of user behavior. But kick us off with just a general high-level introduction on you and the company, and then we'll go from there. Cool. Yeah, it's great to be here. I don't know. I'm not so interesting. The company is much more interesting. the way I like to think of it is we are delivering the best, most valuable digital musical experiences to the whole world. And right now the things that are available to the end user are not always amazing. They're not always differentiated. And I think that's just like a failure of imagination that you can do so much more with music and you can enjoy it so much more.

1:30:40So that's how I think about us. What was the first song that you ever made? Ever made? I've been playing piano since I'm four. I played in a lot of bands in high school and in college. I play every day. So I don't actually know the answer to that, but it's probably like chopsticks or something. Are you a foundation model company? Chopstick is a classic. Or an application layer company or both? Look, the answer is both. All the technologies are ours. These didn't exist. But I think about, and maybe this will color the conversation, and certainly when you guys were discussing us last week or two weeks ago, all that matters here is the product.

1:31:16All that matters is that we deliver an experience to a user that makes them feel a certain way. And so at the end of the day, if there were a way to do what we're doing without AI, we would probably do that. There just isn't. Yeah. Yeah. So there is a win condition where you don't even train models and you use someone else's models. But as long as you have the best application, that's where the value accrues. I think so. Yeah, that's kind of a challenging story. Yeah. Users don't care. And I think probably at some point, maybe there's like a last model we officially release. Yeah, yeah. And the rest is just like amazing product updates.

1:31:53No, no, I completely agree with that. That makes sense. Yeah, that makes a lot of sense. Walk us through, we normally don't talk about the history of companies too much during interviews. But because you're, I mean, this is the first time on the show, like I'd love to kind of understand the different inflection points for the business. I mean, you guys have just grown tremendously over the last year. So walk us through kind of the history of Suno and then kind of the history of even AI music and how those two things track. Sure. We launched this product a little over two years ago. I think it was in September.

1:32:31And it was kind of barely passable back then. It was a Discord bot. It's actually something I got wrong completely. thankfully i have good co-founders um who disabused me of my bad ideas and was that like a mid was that a mid journey were you inspired by mid journey yeah it was working for mid journey why not it was working so well for mid journey and i thought we'd be on discord forever and then um uh it was like around thanksgiving we released of that year we released like a really thin web app um it didn't even have the full functionality of the discord bot and five days for 90 of the traffic to move over to the web it's like so if ever there's market signal that i got something wrong that was like very very strong um and i think slowly but surely you know we keep releasing new products new models the product gets more engaging the music gets more impactful and slowly but surely i i would say the cartoon you should have is um it becomes more and more appealing to a larger and larger group of people every time you make the product better and so um it's basically been a lot of iteration yes there are inflection points um everything consumer is like a little bit lumpy, but that's the history of it.

1:33:36Yeah. What did the early cohort of users look like? How were they using the product? Why were they? I'm assuming you had a cult following early. What did that early cohort look like? And is it that much different than today, or is it just scaled up? It's very different today. So the early cohort, I would describe as like very tech forward and music curious. And it's like, here's this new thing. You kind of have to suspend disbelief. You need to be really forgiving because the music actually isn't all that good. And as you make the product better, you grow out from there. And now it's like people who love music who are like a little bit tech curious.

1:34:20And actually, at this point, you don't even need to be all that tech curious. There's like a fairly easy to use experience on both web and mobile that you can just kind to dive into. And so I think the early users are indeed like, you know, if you were on Discord, let me put it this way, if you were on Discord, like you are a very different kind of user than the type of person who might find a mobile app. Yeah, exactly. Exactly. Yeah. And so today, I think like part of the debate that we were having, and you can deliver some truth around this is like what like what are kind of the key cohorts that are using and loving suno everyone by now has probably seen you know viral short form videos where people are combining two different songs and and some of the reach on on these assets is insane you'll see you know somebody combining like uh rap artists like future with uh jazz music and and you know getting you know hundreds or tens of millions or potentially even more views on this type of content.

1:35:22But then we've also heard of stories of professional musicians using this to accelerate their workflows. And then we've now heard stories of people that are just creating music just for personal consumption. And so what are the different kind of cohorts look like and what kind of user are you guys most focused on? so the professional content creator be their professional music creator or other content creator this is like mid single digit percentage of our user base and the vast majority of people are people um who love music but aren't making music uh for any other purpose that's just we call it creative entertainment this is just like a new behavior and so i think trying to apply too much of what you know from other things be it consumer social platforms or video games it's always you always just have to take everything with a grain of salt um but most people they come and they find that music making is incredibly enjoyable and um if you have studied an instrument and made music with your friends like you probably already know this and um you just don't believe that it can be done digitally and then you kind of find the product and you realize oh my god this can be done digitally i'm gonna enjoy the hell out of making music um i usually um try to cartoonize our average user as people who would say, I love music.

1:36:38I love to sing. You don't love it when I sing. That's kind of our sweet spot. That makes sense. And that tracks on actual monetization. It's not like those are the free users and then all the revenue comes from professional people or API. It actually matches roughly what you described. That's right. You guys mentioned a couple of weeks ago, the Chris Dixon thing, the comfort of the tool, stay for the network. I think about it a little differently for us. I kind of think of it for us as like, come for the gimmick and stay for the joy. And this is like the party trick of making that song that even if it's not, you know, your favorite artist, you know, which largely can't platform, but it's, you know, make the country song about Debbie being late.

1:37:23That's the party trick. But there's something that is behind the party trick that you get pulled into. And all of a sudden you are a music maker and you're enjoying it. And so I actually, I don't think it's bad that um you know there's like this this this song that is meant to be consumed once that you're never gonna consume again that like you laugh for four seconds with your friend i think we know one of you said like i almost just need to know the prompt um and then i don't need to know the rest of the song like i don't actually think that's bad because there's this amazing experience that it gets backed up with that people spend hours a day on totally yeah and i i when i think about some of the most fun moments of my childhood it was my dad writing and playing a song for guitar and singing it one time and never singing it again.

1:38:06And I still just remember the moments and the joy of that. And so the concept of ephemeral songs that can be created for a moment that historically you would have needed studio time, you would have needed to spend all this time writing the song, producing all this stuff, and now it can be done in an instant. I think it really is totally magical. I think this is similar to this. There's ephemeral pictures on your phone. Right. And you never go look at them again unless Apple tells you to. Yeah. Yeah. Where, where are the rough edges or where, where are the edges that you want to be like careful around?

1:38:41I feel like, uh, with chat GPT, no one was really expecting the whole, like getting one shotted and like psychosis and like delusions of grandeur. Like that came out of nowhere in my opinion. And then all of a sudden it was like an issue and they had to work on it. And of course there's like guard rails that they can put up, uh, mid journey. I think a lot of people use that as art therapy. I haven't seen that many people or heard that many anecdotes about people like going crazy because of mid journey, but they're still like, if you're doing image generation, you don't want to generate adult content, for example.

1:39:11Do you have an idea of like the shape of the battle that you're going to be waging against like the risks that you want to like fight off? Yeah, look, I think for us, they're actually just fairly lower stakes than those other ones. Like our AI is turning anyone into paperclips and not that I think that that's a real thing anyway. And like you said, I don't think we're inducing psychosis in anyone. You know, this is a tricky question. Music, the content has kind of always been filthy in the best music. And I don't really want us to be the arbiter of that. So we have like some content moderation that's like pretty, I would say pretty forgiving.

1:39:53And currently that's actually, that's it. And then lots of copyright moderation, of course. Yeah, I was raised on Little John and the Eastside Boys. And it was indeed filthy. Yeah. And yes, it would not be your place to necessarily filter that. I would imagine that you at some point have to do some sort of like KYC age verification or even something that's just like parental controls, that type of stuff. That certainly makes sense. I have a follow-up question. Go for it. Yeah, there was an article in the journal yesterday about how Instagram had used PG-13 as a heuristic for the type of content that they would show to teenagers on Instagram.

1:40:37And the MPAA actually pushed back against that. And I was very upset about that because I feel like we need shared language for how we describe what is acceptable on new tech platforms. so like i want you to be able to put the label of like explicit content remember that explicit content on every album it became so iconic it was actually cool which maybe had a bad effect but if you put that if you put that black and white flag with the explicit on there i know exactly what i'm getting and i know that i'm going to keep that out of my hand out of the hands of my kids for a number of years um and i would hate for i would hate to be in a situation where you couldn't use that.

1:41:18And you had to come up with some new jargon for your policy. I love just building on the shoulders of giants, but have you thought about adapting any of those terminologies or just making your moderation language match what people already know? Is that relevant to you? Not as much as it is on other platforms. I mean, but it is, you know, I feel like we should not be the arbiters of a lot of these things. And, you know, quite frankly, maybe like you don't want us to be the arbiters of these things and that there should be other guidelines set by other things. And so like, yes, if you're dealing with children, you're going to have different rules about the songs that you can make and that you cannot make.

1:41:59And actually, right now, parents making songs with their kids is actually a mega use case. It's probably half of my usage, which is a lot. I use the product a lot. And so it's like right now this is actually not the biggest focus for us um like what we have like works pretty well there are always going to be edge cases there are edge cases in moderation on every single platform literally you know ever and of course kind of just through them as they come yeah what are your conversations like with people in the music industry there were there was some news earlier this week around the first like ai artists signed a record deal i don't know if it was actually the first, but that's how it was being reported.

1:42:38How are the conversations with are going with your users that are like true pro users that are using Suno to either like, you know, experiment with like a bunch of different variations of songs, get new ideas, make entire songs themselves. What are those conversations like? You know, I think there's been a huge shift in the last six months, I would say like now, I really don't meet a lot of professionals who don't use Suno at least a little bit. Like songwriters use this a ton. We run a lot of camps. It's cliche, but a creative companion or a creative co-pilot for helping craft the right songs, you know, verse by verse or getting ideas.

1:43:18Basically every producer I know uses Suno. It was recently put to me just in an incredibly pithy way of we've become the ozempic of the music industry. Everybody's on it, but nobody wants to talk about it. and while that might be frustrating for us that in the interim nobody wants to talk about it in the long run that's actually great you know in the long run the fact that everybody is using it and loves it is fantastic and so it's this weird dynamic I think where one-on-one people are quite pro even though like the industry as a whole may have different feelings about it. It's interesting yeah I wonder I wonder how this like like the creative the creative element of this really throws me because on the one hand, I see if I'm thinking about Suno in the creative realm as like, you know, cursor for your DAW, your workstation, your Fruity Loops or your Pro Tools or your Ableton and you're generating samples and actually working through that, I feel like that is something that, yes, we'd get to 100 % penetration.

1:44:24But on the flip side, Like we've had the ability to generate blog posts and text for years now at human touring test level. And yet like it does not feel like we're close to just, oh yeah, just put in a prompt, write a good blog post and it comes out and it's good. Like you still need that spark, that inspiration. And then most people I feel like are still not, they're still in the co-pilot era for sure. and I feel like that's interesting. Yeah, I don't know. I think there's a big difference though here, which is that there's a few big differences. But for me, hearing you say that, it makes me think a lot of, you're thinking of us largely as a tool.

1:45:09It's because a lot of AI products are tools. And the point is not to just churn this blog post out. Actually, the point is to have enjoyed the process of it. I happen not to really enjoy the process of writing so much. Maybe you guys are different. But for us, the way people use Suno, it's like you're coming. yes, there is some song at the end of it that you listen to and the fruits of your labor are enjoyable, but the journey of doing it is actually why you're there. And so the actual tool analogy doesn't really work there of like, I don't just write in the prompt and it comes out perfect. That's not even the point here.

1:45:39Yes. So I think where I'm getting is that the actual market for AI audio is going to bifurcate, in my opinion, and there will be people Just like there are people right now that use LLMs to just chat and just have an AI therapist, or they have a friend, or they just ask ChatGPT 4.0 for, you know, hey, what should I do at work about this? They're just having a conversation. And then out of text generation, out of LLMs, we also got just the ability to do code completion. And so we got, you know, Windsurf and Cursor and Codex. And, you know, we got these, like, coding agents. And then we also got like AI therapists.

1:46:24And these are two wildly different things from the same underlying technology. And I feel like I would just be shocked if you don't see both emerge over time. I mean, maybe the different company comes in and takes one more seriously than you because if the market's really big. But I would imagine that you wind up servicing both to some degree, no? For sure. I mean, we already service both. But I almost think of this as in the coding example. you might vibe code something and that's just a few prompts or you might actually need fine grained control over the thing and then you use a different tool like something that's basically autocomplete on steroids.

1:47:00I don't mean that in a jarred way. It's amazing. Music is the same way. I might vibe create something and have a session where music is generally what I'm looking for and I enjoy the process of getting there but maybe that's the end of it and if this is going to be a song that is everlasting and intended to top the charts, I need that fine-grained control and I need a different paradigm and a different set of workflows and tools for doing it. And so there may even be the same underlying model that powers both of those things. And it's a different end-user application. Yeah. How do you think video creation is distinctly different than music creation?

1:47:41Because I think, you know, on our side, watching the Sora launch, Sora was like a cool creative tool that had this novel like feature that was the cameo feature. But then immediately those outputs just wanted to go live elsewhere, right? They wanted to go where these big audiences were. And I feel like with Suno, the difference is like you can create music and you want to save stuff and it's actually can be very enjoyable to just listen to the content. Um, is that, is that just a, like, do you think that will, do you think that will always be kind of the dynamic or do you think eventually like video models will catch up to the point where like people are generating videos and then just like consuming them a lot?

1:48:25Because I haven't, again, like, I think a lot of the output of something like a Sora has been like very ephemeral. It gets maybe shared in a group chat or gets shared on another social media platform, but people are not like revisiting the content. Yeah, I think, look, very fast moving space. And so, you know, this is just one guy's opinion today. I think there's basically two big differences. One is the joy in making the music. And I think people mostly enjoy consuming the Sora outputs and the experience of making it is less interesting. And then the other thing about video in general, but especially AI video is it really tends to short form.

1:49:04And actually the beauty of music is that it really is meant to grab you for minutes at a time and not seconds at a time. And it makes for a completely different consumptive experience. And so, and I think that's actually a lot of what's broken in music consumption today is there's so much pressure to put on a song and then put, you know, through headphones and then put your phone in your pocket and not even really pay attention to it. Um, but I, I think a lot more is possible where I do have somebody's attention for three minutes, four minutes, and you can actually tell the story through music in a way that like a 30 second video really does not.

1:49:42And so maybe AI video will get there. Maybe it will get there in a year. Maybe it'll get there in five years. But like today, that's what I see as the two big differences. Makes sense. How do you think the music streaming platforms will navigate AI music? I think they look they already are there's AI music on these platforms uh Spotify's announced that they'll start doing AI music and so um I I very strongly suspect that in like five years this is not a conversation that's being had because I already know that there's little bits of our music everywhere in you know mega smashes and um the same way you don't think about like was this song digitally produced or was it produced on a giant console in a studio somewhere and that's maybe something that people thought for five years you know 30 years ago i my very strong suspicion or just like auto-tune where now every song has auto-tune in it even if you can't hear it my very strong suspicion is that that is where everything um ends up and the reason is it is better for consumers it would really be bad for consumers and therefore bad for the size of the ecosystem if these things kind of bifurcate.

1:50:52Yep, that makes a lot of sense. It's like asking, I don't even think anyone ever asked the question, how will social media apps handle Photoshop? Photoshop was, Photoshop outputs were just immediately, the tool existed prior to social media, so it just proliferated immediately. I mean, speaking of Photoshop, Photoshop has had to grapple with the existence of generative images through partnerships, through rolling their own. How important is it to you to play in the market of like the enterprise audio workflows, do deals, partnerships? Because what we saw in the text, at least in code at least, every coding platform, and there were a whole bunch of new ones, like there's been a real battle for the IDE.

1:51:42And I'm wondering if there's like a battle coming for the top end of the music production workflow. I think there probably will be a battle, but it will be not nearly the bloodbath that it will be in other domains. And it's just because the average piece of music today is produced in 800 different tools. And everybody has all of the tools and so i actually don't think it's entirely reasonable to think that there will be one tool that uses up all of your screen time and like that's going to be the thing that you produce 99 of your music in if you are a true professional and so there um the the end goal is the perfect song and and a little bit less the journey of how you got there yeah interesting code it's like vs code is just open all the time yeah and i don't want to have another editor and if you like talk to people who develop iOS, they're pissed off because they have two editors.

1:52:36But it's not nearly the same 800 different tools that it is in music. Yeah, no, that makes a ton of sense. Makes sense. This is super interesting. Thank you for coming on to break it down. Yeah, this was awesome. Thanks so much. I'm going to use Suno to try to make bedtime with my three-and-a-half-year-old a little bit easier later tonight. So I'll report back. I'm excited to hear it. What are some of your favorite prompts that you use with your kid or kids? You know, a lot of them are songs. I have a one-year-old and a four-year-old. It's still too early for the one-year-old. A lot of them are songs about him in fantastical situations.

1:53:16My most favorite one is just last winter after the Zamboni driver at the local ice rink let him ride in the Zamboni. It was like a whole album of songs about him riding in a Zamboni doing all sorts of crazy stuff. Um, and yeah, they're like very special and personal. And I do go back to those and, and, um, gosh, yeah, I think this is wonderful. Sometimes we joke if we were a kid's app, we'd have five times the revenue that we do. You know, it's just like, um, kids is like the most straightforward way to get me to part with my money. I'm sure you're the same way. Totally. Totally agree. Suno kids, maybe the iPad, the iPad app.

1:53:52Yeah. Coming next year. Awesome. Not actually. So great to meet you, Mikey. Thank you for coming on. And congrats on all the progress. We'll talk to you soon. Bye. Let me tell you about numeral.com. Sales tax on autopilot. Spend less than five minutes per month on sales tax compliance. Speaking of kids. So nice to drop the HQ. Yeah, drop the HQ. RIP to numeral HQ. Speaking of kids, I want your review of this way to decide how to buy a house. So in New York City in 2023, this real estate agent worked with a couple who wanted to purchase an apartment on the Upper West Side of Manhattan. But they wanted their Pokemon-obsessed six-year-old son to have a say.

1:54:39How? By playing Pokemon in each neighborhood before his parents would commit to a purchase. We spent an afternoon covering four neighborhoods as the child played the game and caught Pokemon characters. It turns out that Central Park West was a hotspot for Pokemon Go characters that day. And we toured a two-bedroom co-op that was listed for about$2 million. The child ended up catching a Snorlax as soon as we stepped inside Central Park. A Snorlax is a blue-green character that looks like a bear. It was apparently extremely rare to find because it spends the majority of its life asleep. The child is screaming, jumping up and down.

1:55:18He was so excited to find it. Although my clients looked at seven other apartments that day, no other area came close in terms of Pokemon action. So they knew that apartment was the one. They wrote up an offer that same day for the full asking price. What's your review? Is that a good way? Is that cute and adorable to bring your kids in? I think it's cute. Or is it weird? It potentially could lead to total heartbreak. You know, if the kid's just like, yeah, this is the Pokemon. It's actually a dry spot now. This is a Pokemon hot spot, and then the Pokemon dry up a week in after moving in. Good take.

1:55:52And they're like, what are we doing here? I know we just moved in here, but we got to leave. We got to get closer to the Pokemon. Mom and Dad, we got to book a wander. We got to find our happy place, find our Pokemon hunting spot. Book a wander with inspiring views. Hotel-grade amenities, dream beds, top-tier cleaning, and 24-7 concierge service. It's a vacation home, but better, folks. Yeah, I don't know. I think I like the idea of bringing the kid, the six-year-old, into the house purchase process. But let's leave the phones at home. And let's develop a taste for, let's become snobs. How about that, six-year-old?

1:56:30Become a snob and say, no, I couldn't live anywhere, but. I don't like the cafe. The Upper East Side, because I have strong opinions about this. Get into the lore. What happened on this street? That's what a six-year-old should do. I'm fine to delegate to the six-year-old, but not to Niantic, the company behind Pokemon Go. Anyway, our next guest is waiting for us in the Restream waiting room. We have Ahmad from Mercury. How you doing? Good to see you. Welcome to the show. Great to see you. Hey, good to be back. How are you doing? Great. It's been a little bit. Yeah, it's been a while. A little bit.

1:57:01Great to be back. Give us an update on what's going on in your world. And then I want to just talk about all the crazy stuff that's happening in tech, honestly. Yeah, let's go for it. We just announced today on Fortune Termsheet that we had three years of profitability. Whoa, whoa, whoa. I'm excited about it. We're hitting all the sound effects. Founder mode. That is wild. Yeah, full founder mode. Yes, it's been a - One weird trick. VCs hate this. VCs hate this. One weird trick. Area man discovers profitability. Congratulations. It's actually really funny. I love my VCs, but there's definitely every board meeting.

1:57:44They're like, okay, how can we spend this money a little bit faster? I was actually talking to some CEOs just last week. Most of the time, spending more money doesn't mean you grow faster. And that's kind of what I was trying to celebrate with this milestone. We're building software companies here. At least we are, but not everyone. You don't need to spend money to grow. I got a pitch for you. You're in a strong financial position. you should put out a press release and say, we don't need a backstop. We don't need a government backstop. We are profitable. I didn't know there was a backstop available to me.

1:58:22I was ready to use, I was ready when those comments started floating Wednesday, I was ready to use the analogy of like, what would happen if there was a government backstop on venture-backed corporate cards? I mean, truly, you are a bank and you've obviously lived through, you lived through the financial crisis. I can't say he's a bank. No, okay, okay. We're not a bank. We're a partner with banks. A partner, but you're obviously intimately familiar with the financial system. Do you have any unique lessons or stories that you tell from the financial crisis, warning signs that you pull from, even just like movies or books or something that you keep going back to as like, you know, we want to stay sharp on this?

1:59:05I think building trust in what Mercury does It's actually probably the hardest thing we do Because we're asking a lot from people We're storing all their money for their business I had this conversation when we first launched Mercury With someone who'd raised$10 million And he was like, this is more money than I've ever seen in my whole life I'm not going to give it to a startup banking service And now we have customers that have more than$100 million with Mercury So building that trust is always so important And, you know, obviously, I don't know which financial crisis you were referring to. I was thinking of 2008, actually.

1:59:39But I do want to talk about SVB as well. The one that I think about is like the SVB crisis. Yeah, of course. Take us through the SVB crisis. Yeah. What were the biggest lessons from the SVB crisis? Yeah, I mean, the biggest thing for us was it can't just be all talk, right? And I was talking to founders saying, hey, you should trust Mercury. And they'd be like, oh man, SVB is a 40-year-old company and they just are dying. Why should we trust Mercury? And then what we did and what we've continued to invest in is to make it so it's not just about trusting Mercury. It's about having products that deliver trust.

2:00:20So we have accelerated FDIC insurance, which is, I guess on the website we publish 5 million. But it's actually, most accounts give way more than 5 million. So almost all the deposits that sit at Mercury are FDIC insured. And that was the big thing. I think SVB was at 80-ish percent uninsured deposits. And that's just because, just for anybody that's not super familiar, that's because they had standard$200.25 million of FDIC insurance, but almost every account had well beyond that because they were startups. Yeah, especially VC-backed businesses. And also in a zero-interest rate environment. In a zero interest rate environment, you're like, I'm just going to keep it in the checking account.

2:01:01I raised a$5 million seat. I'll just literally keep it in the checking account because what's the point? If I move it to some other vehicle, I'm going to earn 0.5 % or 0.1%. Now it's a different environment. Yeah, was it always a plan to get profitable, stay profitable as fast as sort of reasonably possible? Or did that happen because of SVB? If people are trusting you as a financial platform institution, it's not exactly comforting to know that you're burning money and your financial institution. Because I think the other thing that was part of SBB was people just had so little time to move all of their banking operations.

2:01:35And it's just like there's so much reliant on a bank account for a business. It was such a funny moment. I was with SBB. It collapsed. And I was like, I need to go someplace less risky. I'm going to First Republic. And then that collapsed too. But that feels like an even older – I don't know which one's actually older. I think First Republic was older. but it was this odd thing. I think Besserit was a little young, I believe. Total refutation of the Lindy principle. It was like, actually go to Mercury. Would have been better off. Sorry, I got it off. I was going to say, I think one of the parts of that was like, you realize so much of operations and it's even more extreme for Mercury customers because we do more than banking.

2:02:16So we do people's credit cards and invoicing and we've been building more and more features and our customers are reliant on a broader set of features. So I think part of being profitable and sustaining profitability is making that promise to our customers that this is something that's going to stick around. So if they use us, it's not like one day we don't raise a funding round or whatever, there's a crash or something and people have to worry about that. So it's definitely a core part of that, having that trusted service. Where are you guys getting the most leverage from AI? Yeah, good question.

2:02:52The biggest thing we've done is back office stuff. To run a fintech, especially on the banking side, there's just a ton of people that have to do a lot of back office things. You have to upload a formation document and someone has to review it. All of these things are very ripe for AI to just smooth it out, either automate 100 % or automate 80 % so that a human can just review the answers. So yeah, lots of stuff on compliance risk back office has been our biggest investment. We've got lots of, I guess a lot of what we do at Mercury is not kind of follow fads or whatever. So we've done a bunch of things that are user-facing, but we try to add little bits of value here and there.

2:03:38So yeah, we have a chatbot so you can ask your questions and that's been actually, actually it's sort of like 40 % of tickets. Customer questions are just like, hey, how long does it take for a while to settle in? that, you know, those kinds of questions. And that's just a better customer experience. And, you know, we think about. And that's 40 % over the baseline of like, because most customer support systems, like if you email in, it would just previously do like fuzzy database search to be like, oh, it seems like you're asking about this FAQ. You got a 40 % lift on top of that because you were already doing best practices, I imagine.

2:04:15Something like that? Yes. It's a bit more complicated because we still have that email system when people email it in. So this did replace the suggested topics in the live messaging system. So yeah, I don't know exactly. What about on the cost side? The suggested answers was way, way less because people don't like reading those suggested things. They just want an answer and they want to move on. What about on the cost side? We were talking to Ivan Notion. He said that his inference bill was actually having an effect on his gross margins. You've remained profitable, but if I'm looking deeper in your financials, will I see an uptick in inference bills based on actually, is that something that you're actually trying to manage as you stay profitable or has it been low enough that it just, yeah.

2:05:05And I think it's wildly different. I think using it in back, back office context where you're sort of processing forms and PDFs and things like that versus Ivan, where people are like, I'm going to generate, you know, trillions of tokens because I'm just making long documents. On the back office side, it's mostly been a cost saving. You're giving leverage to humans. Or you're moving off of a mechanical Turk or scale or something like that. On the user-facing side, I think a big difference for us is we have 200 ,000 plus customers, but it's not that many customers. A lot of consumer services have millions of customers.

2:05:45Like, each of our customers is obviously, like, much more valuable. And also, you probably don't have, like, a ton of DA. I mean, you probably have a lot of DAUs, but it's not the app that people open when they wake up. Yeah. You know? I mean, it's much more utility than productivity. Like, people are actually kind of happy if they don't need to open. You need to do something. That's why you do it. Exactly. And so the inference is just going to be so much lower than if Instagram is all of a sudden being like, we're going to run an LLM query on every, you know, photo that gets uploaded. But it's like, okay, well, that's a huge amount of inference.

2:06:17We're doing more and more, though, so potentially when you're churning through all these transactions, they'll go up. I don't expect for us. We also make way more on a per-customer basis than, actually, I don't know about Notion, but than most prosumer consumer-type things. So I think it would net out that it would be a smallish percentage. Yeah, that makes a lot of sense. What are you seeing across your angel portfolio? you? You've been investing? I guess for context, I've done about 350 investments in the last 10 years. I don't know if I need to go for that.

2:06:56And I just actually announced earlier this year, I have like a formal angel fund now. It was a 26 million kind of angel fund. I mean, I'm still mostly investing in the same way. I mostly invest because it's just fun to talk to founders. And I guess the same reason you guys run this podcast is like interesting to talk to founders and learn new ideas. It's very AI focused right now, as you can imagine. I think AI is kind of fun to invest in because like every six months what you're investing in like changes a lot, right? Like for a while it was kind of dev tools in AI and now I feel like that's done.

2:07:33So most of what I see right now is kind of deeper vertical kind of B2B applications of AI. Do you think AI gives people the permission to invest in more unique categories? Because basically like there used to be like a whole set of categories that were just kind of like small businesses or industrials or some niche some weird niche thing and they would just be kind of off limits because it'd be like oh no no no like the real money's made in in fintech the real money is made in maybe maybe the ed tech but that one's difficult or ad tech or marketing martech you know it's like there were these easy themes but with ai you can go and and actually put together a pitch around something that's um maybe more or less untouched by tech and actually generate a deck?

2:08:30It's definitely like this extension of the software is eating the world kind of vibe that Mark Andreessen talks about. It's eating a little bit more of the world on the edges. Ten-ish years ago, it was all social and mobile. That's where everything was. Then we started doing... Even Mercury's, ten years ago, was probably off-limits to try to make something that went after the banking market. I remember when I started in 2017, it was like, you know, it sounded weird, right? Like it was not something that was done. I think the thing that AI does is like open up labor, right? Like in robotics, it opens up physical labor.

2:09:08And then in the case of kind of the white collar workers, it opens up like human kind of intellectual labor. And that is like a whole set of categories. So, you know, people, generally VCs hate services firms, right? They'll be like, oh, that's like a services firm. But now services firms are cool. It's human. AI plus human services firms are the thing that a lot of people are investing in, ideally trying to get to fully human only. So fully AI only. But yeah, it's definitely opened up a lot of industries that wouldn't have been interesting from an investment perspective. Do you think people have been overly bearish on SaaS?

2:09:52It's notable that OpenAI uses everything from Salesforce to Slack to Qualtrics. And I imagine you use a lot of, on your business, I don't think anybody's out there saying, I'm going to vibe code a bank. Even if you could somehow build that on a banking as a service platform, it's like the level, the amount of work you need to do to maintain a vibe coded bank on top of, it just doesn't make sense. I think there will be a wave of disruption that will happen in SaaS, but I don't think the answer will be everyone's just making their own end-to-end sales force. That sounds ridiculous to me. I hope someone builds a much better sales force.

2:10:40Maybe their pricing won't be per seat and it will be per sale or whatever, but I think the category will exist. The shape of the category will change, the business model will change. But I don't think the answer is everyone's building their own Salesforce. That just sounds inefficient. It just sounds ridiculous to me. I think actually the incremental new applications people build will just be more applications. If the number of vendors you had was 10 and you're paying a lot, maybe you'll have another 90 internal apps that people just made for small use cases. The amount of software people will use will go up, I think, because of Vibe coding and the ease of coding.

2:11:18but I don't think these big categories are going to just disappear because people can now make it themselves. Do you think companies will go multi-product sooner and are you seeing that across the portfolio if the actual engineering hours are less to go from one product to a second to a third that will just encourage more companies? I think it's going to be industry specific. I think a big story about fintech in the last couple of years has been every fintech going multi-product. I think it's partly because when we started in 2019, no one was building a better new bank. So that's all it took. Our first product was a bank account and it was really good.

2:12:03Now to compete against Mercury, you would have to have bill pay and credit cards and invoicing because that's just what people look for. They look for more of a bundle product. So I think it depends where you're competing. and to some extent it gets a little easier because you can just make products quicker. But making amazing products is never easy. I think the quality and craft and all of that do matter. We try to invest a lot. When we make a new product, it's never easy. It takes a lot to make an amazing product. But when you're very new to a category and almost everything in AI is very new, You can build more quicker and then refine it over time.

2:12:47But it's a good question. I think a lot of the ways people end up competing with incumbents, I've heard a ton of, I don't know why we're picking on Salesforce, but I've heard a lot of how do we beat Salesforce kind of strategies. And a lot of them are like, oh, actually we're going to bundle AISDR, X and Y, and we're going to make it a much more fully featured kind of version of Salesforce. so I think like when you compete against incumbents like doing a bundle strategy sometimes is like a better strategy so like we'll see that play out yeah very cool well congratulations to the whole team three year milestone looking forward to the fourth the fifth sixth and hopefully hundreds of years of mercury profitability you got to get past the SVB mark become the longest standing institution It's funny, they were started like a year before I was born.

2:13:39So I have this exact mental motive. Like they were started 42 years ago. That's amazing. Maybe I'll stop thinking about that when they're going to further down. Very cool. Thanks so much. But yeah, really appreciate you having me on. Yeah. We'll talk to you soon. Congrats to the whole team. Have a good one. Bye. Bye. You heard about the importance of AI agents for customer service. You got to head over to fin.ai, the number one AI agent for customer service, number one in performance benchmarks, Mark's number one in competitive bake-offs. Bake-off world champion. Ranking on G2. And he was also mentioning better CRMs.

2:14:11Why not try out Adio? Customer relationship magic. Adio is the AI native CRM that builds, scales, and grows your company to the next level. Brett Adcock fired back at hype around 1X, presumably. Figure CEO Brett Adcock says his, quote, major competitors are using teleoperation, human operators in their videos and calls it deceiving does will he call out 1x by name will he call out elon by name will he call out unitary by name let's find out by watching this clip his analogy it's like if a self-driving car pulled up and we found out there was some guy a demonstration of a robot doing something that's like not teleopt or um like coded and what you're seeing now in the world is like i would say most major competitor to us at this point are putting out most of their content and updates teleopt from a human and i think it's like perhaps some of the most deceiving things i've ever seen it's like uh if i was driving a self-driving car if a self-driving car pulled up next to us we found out there's some guy from tennessee driving it we'd both be like that's uh that's not what i thought this would look like what you want to see is the The funny thing is that, Tyler, why are you laughing at that?

2:15:26Well, that's like exactly how self-driving cars work right now. Yeah. Teleop is actually maybe critical to the development of full self-driving cars and maybe used by all the top companies all the time. Yes, I mean, it is odd that like... I thought it was, I think it's like a key, it's the key enabler for One X to actually get robots in homes and getting them to a point where they can do valuable work. Yeah. Right? And they're going to be taking a loss on that. They're charging like$2 an hour. But that's going to be a flywheel for them to get more and more training data. Oh, the founder of OneX actually replied.

2:16:12I think Dar. That's the founder, right? No, Dar's not the founder. Oh, he's just the VP. Oh, but he's just the loudest voice on X. So I give him all the credit. He's one of the guys. He created OneX. it would be nothing without him. Anyway, he says, so basically Waymo, there were two operators to every car when Waymo started. It is interesting that, yeah, Waymo has never, that I know of, they've never really like published data on how many operators they have per car. I think that that would actually give people a lot of reassurance around like, okay, these cars are driving around. Are there people overwatching them?

2:16:49And then you just get to a point where you say, okay, yeah, there's 10 ,000 operators for 5 ,000 cars. And then next year, there's 8 ,000 operators for 5 ,000 cars. And eventually, there's 5 ,000 operators for 5 ,000 cars. And then eventually, there's 1 ,000 operators for 5 ,000 cars. And eventually, there's one operator for 5 ,000 cars. But it's just a slow process of getting people up to speed on the new technology. And it just gives people confidence at every turn, much like the confidence that I feel when I go to sleep on my eight sleep how'd you feel how'd you sleep last night jordy uh 88 from me seven hours 38 minutes i want to hear if that soundboard still has my favorite song on it boom you got an 88 i'll give it to you wait do we tie no i got an 83 oh 83 okay rough one um rough one well we have breaking news from the pope he chimed in on the ai bubble some bub talk from the from the papacy a papal bull.

2:17:47Is that a top signal? On the bub talk. No, no. It's not on the bubble. It's not on the bubble. It's not on the bubble. It is from the Pope, though, who I should be following. Why am I not following the Pope? Technological innovation can be a form of participation in the divine act of creation. It carries an ethical and spiritual weight for every design choice expresses a vision of humanity. The church therefore calls all builders of AI to cultivate moral discernment as a fundamental part of their work, to develop systems that reflect justice, solidarity, and a genuine reverence for life. I like that.

2:18:29Good line. Liquidity thinks it's AI-generated slop. Very bearish. No. Yes, he used one em dash. Like, the em dash is not a telltale sign. You can just use one if you're just hammering on the keyboard. Or you can just write a note, and then in the edit, someone can use an em dash. I don't know. Is the Pope one-shotted? We'll never know. I think this is fine. I think this is a nice little call to action. What do you think, Tyler? You put this in here, right? Yeah. I mean, what are you guys doing if the Pope comes out and says, like, I actually don't care about AI margins? What are you doing? I would be extremely bullish on that.

2:19:09What I'm worried about is the Pope coming out and being like, actually, I've discovered some novel physics. I've been chatting. I've been talking to Grok, and I've been talking to Chad G.P.T. for just like thousands of hours. People don't think about the recursiveness in the Bible. Yeah, no, yeah, recursiveness in the Bible. Actually, it's crazy. I'm the Pope of the Catholic Church, but I've been getting into polytheism recently. I went down a rabbit hole, and I think there might be more than one God. I think there's multiple gods. uh ab in the chat says got to get the pope on vanta totally agree 100 i i think it would be the the it'd be interesting if the pope started coming out and said we're exploring taking equity stakes and technology companies like uh fannie mae and freddie mac yeah well we'll see if we gotta get the pope on bezel because your bezel concierge is available now to source your watch on the planet seriously any watch uh yeah the pope already has the the drop top top six-wheel G-Wagon.

2:20:06Is it six wheels? I think so. No way. I always thought it was a rules. No, it is a G-Wagon. But I feel like he has a wild set of cars. No, it's not six wheels. No, it's not six wheels. Six wheels would be so... But it is a one-of-one, and they made an EV. I love it. Speaking of EVs, the Ford F-150 Lightning is potentially going out of production. No way. It is close to failing. The Wall Street Journal has a story on the front page talking about, where is this? Pull that up. The Ford F-150, Ford Ways ending electric F-150 truck. Executives at Ford Motor are in active discussions, active talks. They're in talks.

2:20:55We love active talks. About scrapping the electric version of the F-150. People familiar with the matter said, which would make the unprofitable truck the U.S.'s first major EV casualty. Do you think this is all triggered by TJ Parker? Possible. World-famous venture capitalist and founder himself. He bought a Ford, but he didn't buy an electric one. He bought a Raptor R. Why don't they make the electric one like the Raptor R? they should with an internal combustion engine exactly i feel like if they want people to buy the electric version of the truck they should give it like an exhaust and an engine and a real engine note yeah real engine note when i had my raptor i and it's more cylinders for people anytime i was on a call in the car yeah people would assume that i was driving a sports car because there's so much oh it's insane noise in the cabin it was ridiculous yeah i mean the raptor r has how many cylinders does it have?

2:21:57Is it a V8? It is, right? I thought it was. So it has eight cylinders. The F-150 Lightning had zero cylinders. That could be why it's not selling very well. The Ford Raptor R had 5.2 liters in its engine. The F-150 Lightning had a zero liter engine because it didn't have one. So it makes sense why this has been struggling. And we will see what they wind up doing. We got to change the subject because all this Raptor talks got me thinking about getting one again. Pull the trigger. Well, if you think it's bullish or bearish that Ford is pulling the Raptor. No, they're doubling down on the Raptor.

2:22:40They're pulling the Lightning. Head over to public.com, investing for those that take it seriously, multi-asset investing, industry-leading yields. They're trusted by millions. quadrillion what else is going on in the markets hero lewis says my inbox is full of people breathlessly trying to interpret this and it is one year look at spy it's a slight dip you can interpret this in two seconds it's over it's just completely over the bubble popped now it's ready to start rebuilding it was a good run we can go up from here because the bubble has popped you're welcome yep it was rough we got through it we are now it's almost a weekend we are now past the trough of disillusionment, we're in the plateau of productivity.

2:23:19That's what I see. That's good. And pretty soon, oh, the stocks have already stopped trading. So we're safe now. We're safe. It can't hurt us anymore. The economy can't hurt you on the weekend. There's so many more charts. It's really chart talk time. Consumer sentiment, almost back to its record low. People are not happy with consumer sentiment right now. You see this too, Bank of America, just as of yesterday, just fully recovered from the global financial crisis. Wow. Took 19 years. To the day, and then it's like back in another crisis. Can you imagine? It's like, no, we just dug our way out of the hole.

2:24:00No, fantastic. Famously backstopped by none other than Warren Buffett. Yeah, it's been interesting to follow. Although, I mean, Berkshire Hathaway is green this week. He's got so much cash. He's got so much cash. He's building cash pile. So it's$1.07 trillion. Maybe the least AGI-pilled person in the world. Cash-pilled. Imagine if Berkshire Hathaway just comes in and just acquires like 90 % of OpenAI. He likes those authentic greenbacks. AG, what's the I for? Authentic greenback incentives. I don't know. Chart of the year. Daniel Eath says, is commenting, so NVIDIA put out something on Wednesday.

2:24:52As I've long said, China is nanoseconds behind America and AI. It's vital that America wins by racing ahead and winning developers worldwide. Daniel says, this is an insane statement. America is more than nanoseconds ahead of China and AI. The actual gap is months, and it would be larger, too, if NVIDIA stopped the flow of advanced GPUs to China. So putting him in the truth zone. There was some more information on the Snap Perplexity deal. Snap gets$400 million, which is greater than Perplexity's total revenue. Snap gives nothing except access to... Wait, wait, wait. It's more? What? They're paying more than their revenue?

2:25:36How does that work? Perplexity is taking VC dollars, giving them to Snap for distribution. There's also an equity. All the Snap holders are. Sorry. This is extremely confusing. Let's read this. The deal looks incredibly in Snap's favor. Snap gets$400 million, greater than perplexity total revenue. Now, Signal is saying this is most likely$399 million of perplexity equity to Snap, not cash. So there is a question about how much of it is cash versus, so yeah, the raise VC dollars, give it to Snap is one theory. But it is possible that they're just giving$400 million of equity, like a stock grant.

2:26:22I think it's some split. Yeah. I don't know if they published it. But like Signal's thesis is that it's 399 million of equity and like 1 million of cash. And so it's basically just like we want distribution in Snap and we're willing to give you equity to pay for it because 400 million on 20 billion is 2 % of the company, right? Something like that? Yeah. It's crazy. 2 % of the company. Snap gives nothing except access to an unloved AI chat. Perplexity gets indirect access to zero income teams. Sasha is not a fan of this deal. Spiegel negotiation masterclass. Go back to the conversation we were having, I think Saturday.

2:27:02We had a couple hour drive and we were basically doing a podcast with no mics. But I was saying I was surprised that Snapchat has not figured out a way to just monetize all the capital that a lot of these consumer AI companies have given their massive, massive user base. Well, it's because the numbers are all flipped around. Like, Snap is, I think, like way more mature on all the business metrics and yet worth half perplexity. Right? Like, look at the difference in terms of like business metrics, DAUs, page views. Perplexity does not want, will not want people comping it to Snap in their next financing.

2:27:48Yeah. It's odd. I don't know. It's some sort of dance to actually integrate these two things. Tane says, Snap will be integrating Perplexity directly into Snapchat's chat interface. Perplexity will pay Snap$400 million over a year in a mix of cash and stock as part of the deal and gets access to Snap's 900 million monthly active users. ab in the chat snapchat's gonna accidentally build agi just trying to make the dog filter blink realistically i love it do you think tyler's had a pathway to agi well i mean look spiegel i mean spiegel might be back you think so i mean look we know someone who was previously what do you think he actually what do you think he actually gets what do you think he actually gets I have a lot of trouble understanding what perplexity is because it's somewhat of an application layer company, but a lot of the APIs, a lot of the foundation model products do deep research.

2:28:55Some of them have news hooks already. So if my choice, like Apple could have gone to perplexity. They went to Gemini. Does Gemini give you search results the way perplexity? Like perplexity's initial pitch was it's a LLM, but there's no cutoff and we'll search the web for you. I feel like most of the LLMs just do that out of the box now. And so like, what are you getting when you partner with perplexity over something else? What's the differentiator? You're getting cash. That's why I think a lot of this deal had to have been cash because I think if I'm Evan Spiegel and I'm looking at like perplexity's revenue and user base relative to his own.

2:29:34Yes. and you're not okay so help me understand this math if i'm if i'm apple how many how many da used does apple have is it it's like the same size as snapchat right in terms of like actual iphone users it's got to be a billion right there's like a billion iphone users something like that I think it's more. Like 1.5. Okay. And then Snapchat has like 400. Oh, wait, no. Snapchat has almost a billion MAUs, but DAUs is like 400, right? Yeah, like 470. Okay. So let's say that Snapchat has roughly 40 % as many users as Apple. So it's 40 % to Apple's billion plus, 1.5 billion. uh apple is paying google 1 billion snapchat could potentially go and pay google for access to gemini 40 as much right because they're 40 the size so they could pay 400 million and get the exact same quality of of ai product that apple is vending you'd think because apple says hey or google goes to Apple and says, hey, you want to service a billion users?

2:30:53That'll be$1 billion. Oh, you want to service 400 million users? That'll be$400 million because it's basically a dollar a user. I don't know. Google's basically charging a dollar a user, but perplexity is in the opposite scenario where they're paying for the right as opposed to the other way around. Yeah. This is the flip. The definition of perplexity is inability to deal with or understand something complicated or unaccountable. It's perplexing. It is perplexing. Similar to confusion, bewilderment, puzzlement. Yeah. I mean, this was a little bit of my question with Mark Gurman. He was like, no, no, no.

2:31:26And he kept shutting it down. And I think he has a strong opinion and he might wind up being right. But my take was when a platform, when an aggregator partners with a foundation model lab, which direction should the money flow? And he was like, obviously it should flow from the application layer to the foundation model app. So obviously Apple should pay Google because Google is inventing a technology and then selling that technology to Apple. I think, again, it would have looked a lot different. It would have been flipped if Apple was just, if Gemini was becoming the default assistant within Apple, which they would never probably allow, right?

2:32:08Well, they would if they're getting paid. Potentially if they were getting paid$20 billion. Exactly. which is like how I think this plays out actually. I think somebody, maybe it's OpenAI, but I think somebody will be paying Apple for the right to funnel those LLM queries wherever they want and take the cut just like what happens in search. But that's a hot take and I respect people that don't agree with me. In other news, Deutsche Bank is exploring ways to hedge his exposure to AI to data centers. It's looking at options including shorting a basket of AI-related stocks and buying default protection via synthetic risk transfers.

2:32:45So they're hedgemen. So they bet big on data center financing. The scale of the expenditure on AI infrastructure has prompted concerns that a bubble is forming with some likening the enthusiasm to that which preceded the dot-com crash. Skeptics have pointed out that billions of dollars have been deployed in an untested industry with assets that quickly depreciate in value due to rapid change in technology. Stream Yimby in the comments says, if there's one thing Deutsche Bank is really good at, it's being very exposed. So in recent months, Deutsche Bank has provided debt financing to Swedish group Eco Data Center, as well as the Canadian company 5C, who together raised more than$1 billion to fuel their expansion.

2:33:26The investment bank does not break down how much money it has lent to the sector, but it's estimated to be in the billions of dollars. Hedging exposure to the industry could prove difficult because betting against a basket of AI-related stocks in a booming market will be expensive. Meanwhile, SRT transactions require a diversified pool of loans to earn a rating and investors to likely demand higher premiums. Hyperscaler's pursuit of superintelligence has fueled demands for infrastructure that will help them build it. I mean, this is everyone from the Haunted Mansion proprietor. Like, everyone is getting in on the action.

2:34:01Like, a few years ago, it was like, if you have anything that's related to AI, Like go, go, go, raise money, grow it, spin it, flip it, turn it around, pivot it, whatever you want to do. Europe is expecting a wave of deal-making consolidation and digital infrastructure. And the cost estimate from the Financial Times on the total AI build-out,$3 trillion. $3 trillion. Pretty wild times. uh business insider has a scoop here that google is looking to invest in anthropic at a really 350 billion dollar valuation um it's interesting anthropic seems what one quarter the revenue of open ai and yet from a vibe perspective from a from a public attention from a pr crisis perspective anthropic used to be in trouble every other week because daria would be out there being like paperclip, we're going to paperclip everyone, you know, we're going to get rid of everyone's job.

2:35:05And it was like really aggressive. Right. Uh, and so like, like they had to back stuff up, but they had to backstop, they had to backstop their comms a couple of times, but it's been great. And Dario's looked great for the last six months. And he's just been beating the drum on, Hey, there's a, there's geopolitical considerations here with China, but overall we're, we're building and we're happy to be an API. We're happy to be, we're, we're happy to help businesses. Yeah. Generate code. Running, running the experiment of like highly focused lab versus highly, not distracted, but taking a lot of shots on goal with open AI.

2:35:44We could, we could be looking back on this in a couple of years and being like, man, they really missed erotica. They really missed a slot. Consumer electronics. They missed consumer electronics. It's possible. But on the flip side, it does seem like they've built a bit of an elegant business, and it seems to align well with Google's strategy. And a very hyperscaler wants a piece of it. Yeah. I mean, similar to how Microsoft has access to open AI models on Azure, Google, bringing Anthropic in through an increased investment. Also, I just wonder, is it hard to do a deal at 350B these days in the venture community?

2:36:20There was already that reporting that Dario had to go all over the world to scrap together the last round. And I'm wondering if there's a world where if you're trying to raise$40 billion, you should just go to the hyperscalers. Well, yeah, they're using every possible capital source. We've seen some of these SPVs that have like, you know, 0 % carry, 10 % management fee up front. So they're tapping everything. Let me tell you about adquick.com. Out of home advertising, easy and measurable. Say goodbye to the headaches of out of home advertising. Only Adquick combines technology, out of home expertise, and data to enable efficient, seamless ad buying across the globe.

2:36:51Jordy, where'd you want to go next? I think we should bring on our next guest, our first and only in-person guest. Let's do it. Jordan Castro. A day. He's here live in the TBP in Ultra Dome. And we will bring him in. Muscle Man himself. He is the author of Muscle Man and The Novelist. Wait, was that your first novel? Did you write a novel called The Novelist? Yes. What inspired that? It's kind of like, it's recursive. Yeah, well, I wanted to say, first of all, thanks for having me. Of course, of course. You know, I, yeah, I, and shouts out to Norman Plays in the chat. Someone just showed me, they said, I'm here for Jordan Castro.

2:37:28Oh, let's go. There we go. Shout out to Norman. But yeah, I write novels, but I also, you know, you guys had the Pope up talking about the Pope. And I work with this guy, Luke Burgess, with this thing called the Clooney Institute. Oh, no way. And one of the things I've noticed, it's sort of, we sort of do stuff at the intersection of like what we say, Athens, Jerusalem, and Silicon Valley. And so sort of bringing religious wisdom into the tech conversation. But so I've been like, I'm not really related to tech, but I've been like at all these different like talks and so on about tech and whatever.

2:37:59And I noticed that when people normally talk about AI, they sound like AI. They immediately just sound like AI. And when you guys were doing your AI quadrant thing, I was like, they sound like people. And I was like, why do I like people? Let's give it up for people. Yeah, yeah, let's give it up for people. That's right, that's right, that's right, yeah. And I was like, why do I like TVPN? You know, because I shouldn't like, I mean, you know, I shouldn't like it or whatever. And I was like, well, because you guys like talk like people and you talk about people. You know what I'm saying? And so much of the conversation about tech.

2:38:24Yeah, it was very deliberate to put the faces of the people there, not the logos of the companies. Yeah, and I noticed that they weren't particularly flattering photos. Yeah, yeah. Oh, interesting. You mean we didn't make them in the giga chats? A lot of times we'll give them muscle. Well, we will literally make the muscle man. We have fun with that. But for that, that decision was actually more just about like how do you actually fit them all in? Like if you make them the muscle man, then you have to make it smaller. You can't see their facial features. No, I mean some of those photos are like their hero photo that like their comms department clearly like put on their Wikipedia page or something.

2:38:57I don't know. Hey, man. Or like their corporate photo. You don't always get to pick what photo becomes you on the internet. But sometimes. Trust me. I know. When people get mad at me on Twitter, I've been publishing since I was like 16. Oh, do they go find a stupid photo of you? They find pictures of me in high school wearing a pink button up, and they post it in their comedy. That makes sense. That makes sense, yeah. Yeah, I have some silly photos of old that I'm sure will resurface when people want to dunk on me. Break this post down for me, actually. What was your read on this? So technological innovation can be a form.

2:39:26You think he used AI for it? Do you think he did? Well, I think the worst thing about AI is that people are saying you can't use MDashes anymore. I agree. I love the MDash. Okay, okay. So I don't love the M dash. It's not that I hate it. I'm just indifferent to it. I don't even know where it is on the keyboard. And so you do minus sign, minus sign, and then space, I think, to generate one. Yes, you want me to shift. Double tap. Double tap space. Shift, option, dash. Isn't it just double tap space? And I go here. You can double tap it, but sometimes it doesn't show up. You're like, this is the story.

2:40:00And then I want to do an M dash and then space. Hold shift and you can hit it. Oh, you can hit shift. Putting on an absolute human M dash clinic right now. That's right. That's an underscore. That's an underscore, dude. Look, underscore, underscore. Maybe it's just something that I know how to do. Test, underscore test. How do you do an end up? Maybe, yeah. I just run it double dash space. Maybe it's not good. So I completely agree with you. So I write like a little 300-word, 400-word summary of my current thinking on whatever's in the news every day. I just write it right in Google Docs here.

2:40:34I don't use any AI. I hand it off to Brandon on our team. he edits it, I don't think he uses any AI, sometimes he would put MDashes in, and if we put in even one MDash, all the comments would be, MDash, AI, AI, what was the problem? And it's just like, okay, now I just can't use the MDash because the masses will go crazy. I refuse to let them take that from me. You're fighting the good fight. Yeah, yeah, yeah. You think you can win? I think that it's just, I lost it. And Brandon was like, really? And there were places, I'll show you, sometimes I will literally just write, like a minus sign, not an em dash.

2:41:09I'll just be like, you know, close a quote. I will deliberately break the rules that I know I'm not supposed to. Like even when you put something in quotes, you put like the period inside the quotes or whatever, I will just break all those rules because it just makes it clearer that I'm not slopping it up. I don't know. I think you're accepting the frame of the haters. I think so. I think it might be. I think you're right. I think you're right. Also, shouts out Brandon. I didn't hear any claps for Brandon. Yeah, let's get some claps going for Brandon. We call him Brandon. He's got a new haircut.

2:41:40He got a new haircut. We need a trading card for Brandon's new haircut. He's a sleeper muscle man, too. He's a muscle man. I know why. Is it fair to say I got you into it? What? You're the reason? He's the reason you got jacked? Is that fair to say? He's the reason that you're absolutely shredded and diced. He's the reason you're walking around at 2 % body fat? Okay. Wow. Brandon content. That's why you can bench three plates? You can just wrap it out because of him? Four. Okay, four. Wow. Whoa. Whoa, whoa, whoa. Look. Crazy. That's good. Wait, so yeah, actually tell me the story of the Muscle Man book.

2:42:17Like what inspired it? What was the thesis? It's a novel about an English professor who hates being an English professor and loves lifting weights. Sick. And part of it was that - And this is a real person or - No, it's a fictional character. Okay. Yeah. Yeah. Yeah. But like, you know, part of it was that I when I was growing up, I basically grew up online. And, you know, it's sort of a classic story, but it was like I was totally anxious, you know, like I would be at school. Then I would get on the screen and, you know, I would just post and sort of lurk and be sort of weird and creepy in front of the computer, you know.

2:42:48And then when I started lifting, my anxiety and depression just like completely went away. Yeah. You know, and I was like, damn, like, you know, was my dad right my whole life? You know, are all the sort of jacked guys who are like annoying and saying that, you know, you got to start lifting. Were they just all totally right? The other thing that tracks with my personal experience, I used to have trouble falling asleep at night. Yeah, that too. And the second I started lifting daily or six days a week, it just went away entirely. And the lesson is people that struggle to fall asleep, at least in some cases, are just not exerting themselves enough to be just physically tired enough to like fall asleep.

2:43:30Yeah. Started lifting soon enough. It's like, okay, you basically close your eyes. You're knocked out in like five minutes. Totally. Yeah, yeah, yeah. And so I wanted to just kind of like, and then I tried to find novels or even books about lifting that I thought were like compelling at all. And I couldn't find any. So I started sort of trying to write something like that. It's also like a satire about higher education. There's like a lot of rants and stuff about the universities and stuff. That's cool. So, yeah. Talk about the reception. The reception has been sort of idiotic. Okay. You know, the media, as we all probably.

2:44:03Let's give it up for idiotic. Let's give it up for the idiotic media. Yeah, yeah, yeah. You know, it's. The idiotic media and we have a horn going off. We are the idiotic media. Yeah, yeah, yeah. You know. You have a horse wearing a hat in the background. Idiotic good. We're not journalists, but I definitely consider us to be part of the idiotic media. Yeah, well, now you're using it in a good way. You're reclaiming it, you know what I mean? But most of them are idiotic in a bad way. And the reception has been like, here's a novel to explain the manosphere. The manosphere, right? And here's a novel about men.

2:44:38And I feel like back in the day, there's this great, the literary, the writer Norman Mailer has this debate with these feminists back in the 70s, and he keeps calling them lady critics. This is a very hilarious debate. And I feel like in a weird way they're doing the same thing to me, but they're calling me like a man novelist. Okay. You know what I mean? Where it's like, this is a novel about a man. Sure. So like it has to be about masculinity or something like that. Yeah, everyone wants to know what's going on with young men. Yeah. And it's this kind of fake discourse. Yeah, I saw Chris Williamson on Tucker the other day.

2:45:08Yeah, yeah, exactly. It felt very much, I mean, honestly, I feel like Chris has done a fantastic job of actually sort of explaining what's going on with young men. It is an interesting question. Yeah, yeah. There's a whole bunch of, And I feel like every media outlet has their own way of interrogating this, whether it's Harper's or Vanity Fair or Wired. Like everyone's telling stories and writing profiles about interesting people and young men. And I don't know. It seems like it was somewhat worthwhile discussion to have if it can inform interventions, even just in your own life or with your own kids.

2:45:39It seems like there's some value to the discourse. But did you have any intention of actually sparking that debate? or like actually like like was was one of your thesis like i need to put this book out so that more people lift weights at young ages uh no but i did write an essay for harper that was just basically a straightforward pro lifting essay so that had more of that sure sure sure that's awesome the uh the thing that the my takeaway from going i was very anti-gym growing up because i was a skateboarder and yeah me too right exactly skater surfer like the idea of like going inside being indoors, lifting stuff up and putting it down just for the, whatever the, whatever the reason was, it wasn't appealing to me as a kid.

2:46:23Cause I was like, well, I could just go to the beach and surf and like be out in nature. And so I had a generally negative sentiment around the gym. And I was also like super scrawny growing up. I've got like long arms. I'm, I, I was like in high school, probably like 140 pounds, like getting under like one plate on bench was like, I'm gonna die at first and I never really got into it I wasn't I didn't play football or anything like that uh and then I got really into it in college and it was the biggest unlock for like mental health yeah like I prior to prior to that I was like a you know teenager I was like teenagers I think are just naturally kind of like moody and figuring out who they are and how they fit into the world and all that stuff and then I found lifting and like a lot of that stuff just like went away entirely because I had a purpose.

2:47:12Uh, I would, even if that purpose was like so simple and just like going, I woke up, I was going to eat well and I was going to go to the gym. I was going to lift as heavy as I could and go on with my day. And it's just, it, it ended up being like, uh, it, it was like a drug almost immediately. Cause I was like, I want the feeling of how I feel after I lift. It was, it was incredibly addictive in a very positive way. So I did that. And then about probably a year or so, I went from like 100, by that point, like 150 pounds to like 190 in like a very short period of time. Like basically spending, this was from freshman year to sophomore year.

2:47:50And I was spending like so much of my energy just focused on lifting. Yeah. And I ultimately reached a point where I was like, okay, at some point, this is actually not the most productive use of my time. and I was fortunate enough to at that point to kind of like discover like work and things that I was passionate about and and ways that I could use my energy that was more interesting to me and so I think like the debate to have around lifting is like I think like basically everyone should be like picking up heavy stuff and putting it down like at least a little bit right you don't need to go to the extreme uh the the advice that I find myself maybe giving like you know people that were in my shoes is like figure out the point where you maybe actually want to dial it back a little bit because it's possible and probably healthy to go way too much to the extreme where you're spending like, like I was like all my time and energy, like just lifting because I didn't have anything better to do besides like, you know, work and school.

2:48:48And then at some point I actually need to dial it back and actually apply that energy in other places. But I got the benefit of like that period of just like obsessiveness that I still carry with me. yeah 100 % I mean I that was sort of actually what the novel is exploring you know because when I first found it like you said it's like is this the answer is this the ultimate answer you know what I mean and it it really is I mean that's a part of the issue that I take with like the optimizers you know like Huberman I mean I know you guys had a Huberman on or Brian Johnson where you sort of become like perversely obsessed with health or with lifting or something like that you sort of tinker you know you think of the person the human person as some sort of like mix of neurons and chemicals and stuff.

2:49:28Yeah, it's interesting. Even between Brian Johnson and Hewerman, I see a huge wide gap. There's a big gap, yeah. But let's put them in the same bucket. Yeah. How do you contrast your philosophy? I think you need to be injury maxing. Okay. I think you need to be horsing heavy weights. Interesting, okay. And I think it's like a spiritual endeavor. I think you need to almost, I think you need to get crushed under a body-breaking weight a few times. And, yeah, it's more of a spiritual process than an analytical process. Yeah, 100%. Eric Bugenhagen, who's like my favorite lifter on YouTube, always talks about how it's a mindset.

2:50:10And he has these things where he'll just fail insane amounts of weight multiple times and then just like work himself into this kind of like crazy mental state and then lift it, you know? What do you think about the other YouTubers? What do you think about Sam Sulik? Oh, I love Sam Sulik. What do you think about the Trend Twins? I love the Trend Twins. I love, yeah, yeah, yeah, yeah, yeah, for sure, for sure. Sam Sulik, also Sam Sulik is weirdly emo. Yeah. Like he will, you know, his vlogs, actually there's this weird thing where, so my, I don't know why I'm talking about this now, but my brother died like a year and a half ago.

2:50:40And after the funeral, one of my buddies threw on a Sam Sulik video. and I don't yeah it was incredibly inspirational but I noticed I was like Sam Sulik's videos are sort of emo where he's like driving by himself he's like talking to the camera he's like taking unnecessary loops around the parking lot I think that's why he's so popular I've maybe watched like 20 minutes of Sam Sulik I appreciate like the content like that the world that he's built but because specifically I remember being in that mindset of like I was kind of like again I remember I was like 19 I was in school I wasn't that happy.

2:51:15I was like living in Santa Barbara, which was like paradise. But like mentally, I was chasing that. Like I need to I'd be like that those scenes where you're like driving. I'd be driving to the gym at like 5 a.m. I can relate with everything you're saying, like down to the specifics. Like I started lifting to like sort of like I was in school, just moved. Didn't really have friends. Would like go to the gym at 11 o 'clock, you know, weird hours. Early, early mornings or late and then getting like food afterwards. And like, honestly, that era for me, like future, uh, the, the rap artists was like absolutely peaking.

2:51:48What year was this monster for me? Yeah. so I became, we might, we might be the same guy. I became a man at that time. Is your full name Jordan? Yeah. Wow. This is brother. Make it make sense. Make it make sense. Make it make sense. What's going on here? But, but yeah, so I would, I would drive. I, I just remember like at the time I was driving a, a Prius with like 200 ,000 miles on it. And I'd just be like driving to the gym, just blasting future, drinking protein shakes. And that's like when I think I became a man. yeah like it's incredible incredible things and so i want i want i i think it's super important and like when i hear when i hear that a young person is getting into going to the gym i'm like great totally like you're gonna become you're gonna find yourself through that and my only thing is like i think at some point you have to evolve beyond that and take the learnings and that like that mentality that you get from going to the gym and apply it to other other ways i think like the big takeaway that I learned early on was like early on, I was obsessed with like, how many, how many sets am I doing?

2:52:57How many reps am I doing? How am I, uh, how am I, you know, just approaching different lifts, how many different exercises should I be doing in a lift? And I very quickly realized that there was like two factors, there was maybe three factors. It was like, was I sleeping a lot? Was I eating a lot? And was I like hyper consistent and doing a lot of volume. And those are the same lessons that I've brought into my real life outside of the gym. It's like, am I sleeping a lot? Am I like nourishing myself? Am I getting the right amount of calories? Am I getting, you know, getting the right amount of, um, uh, just like the right kinds of food.

2:53:33And then am I being consistent in the work? And so we apply that every day. We come in and do the show. Uh, and you also go to the gym. I mean, I don't know, a behind the scenes look the whole squad was at the gym this morning which i think is a great thing that you guys do yeah uh a lot of people that were on that chart a lot of the tech elite are deeply unpopular people um i think elon was polling is less popular than donald trump so not just controversial but actually just unpopular across the board a lot of tech ceos a lot of tech people are just unpopular in america obviously we love them here um but we're weirdos about that But my question is, when I think about who's really popular, I think Tom Cruise, I think George Clooney, I think people that are in shape.

2:54:15Is there something, we do this jokey giga-chatification of these people. Is there something where it would actually be in their best interest for the tech elite to get a lot more space? Look at Bezos. Look at Zuck. I was going to say that I used to think Zuck was a full creep. He was like a beady-eyed sort of bug. When Zuck would look in the camera, you could almost hear him blinking. You know what I'm saying? And then he got jacked and threw on a chain. I was like, this guy's not so bad after all. You know what I mean? So maybe they would all benefit from it. PR teams hate this one simple trick.

2:54:54Well, PR teams should love it. No, they should love it. But they maybe aren't awake to it for some reason. Yeah, look, if you're on a billionaire's PR team. Step one. Step one. Trend. Hire a trainer. Bring the trend twins in. Yeah, bring the trend twins in, get jacked, and then the public opinion. Why do you think health is so political? Oh man, I did an event two nights ago here and someone asked the same question. and this guy in the audience stood up and he, because I actually don't know. I mean, one, one reason is that I think that when you're jacked and you're strong and you're, you know, or you're competent, it, it acts as a judge actually, because if you're like, you know, fat and you're lazy, then you feel just sort of implicitly judged.

2:55:43The other thing is that going to the gym and getting jacked implies a certain value system. You know, there's, there's like, you know, if you lift 145, you know, that's less than 225. And if you're in the gym, you're trying to get stronger. The implication is that it's better to lift 225. And so like a lot of people who are like dogmatically committed to a certain kind of like egalitarianism don't like that. But the guy at the event stood up, this guy, he looked just like Joe Rogan. His name's Anthony, actually shouts out Anthony. But he was like, he said that he said that generally speaking, the left tries to impose their ideas onto nature and the right starts with nature and discerns their ideas from there and so there's something just sort of fundamentally different about that i think lifting sort of falls into the latter camp i have a question yeah i'm always i'm always surprised at how much like there seems to be more controversies within in between like various people even that consider themselves all to be pursuing their ideals of health it's like everybody like kind of picks a lane in a category and then it's like they're fighting over like green powders what green powder should you take oh my green powder is better than your green powder it is universal like it i feel like it's it's hard to make a political issue out of something that doesn't affect everyone because you just keep bumping into people and if your whole thing is like you know corporate cards or something or like yeah you know the the manufacturing supply chain for large gongs.

2:57:05It's like, there's just a lot of people that are - You guys have a lot of those, right? Yeah, yeah. You even have the, I noticed in the bathroom, the picture of the jacked guy, the shirtless guy hitting the gong. I think these guys love the gong. Yeah, yeah, yeah. It's like, it's hard to get everyone animated about that. It's a niche interest. Whereas like healthcare, how much money you have, your job, your health, your food, like these things are universal. So anyone can, anyone can relate to them and I can give you a take and you can bounce off it immediately instead of being like, I don't really know anything about that.

2:57:34That doesn't really affect me ever. Yeah, but there's also, there is so much, more people die in this, this is actually, I actually don't know if this is true, but you see this talking point all the time. More people die in this country from like obesity and from starvation. You know, there is this total cope around being unhealthy and people don't like the idea that there's actually something you can do about it. Yeah, yeah, yeah. It's just like everyone can be obese or not. Like not everyone can have a, like a large corporation that goes bankrupt under their stewardship. So it's just less relevant to them because they might not be the CEO of a company that's going bankrupt or something.

2:58:07Pickleball. I couldn't find it in here, but in the mansion section online, there is a house that is up for sale right now that has a pickleball court outside and a pickleball court inside. What's your take on pickleball? Dude, it's fun. I hate to say it, but it's fun. Guilty pleasure? I mean, I played it one time with a friend. And I was like, man, this is – I probably said some words. And then we played it, and I was like, this is awesome. Yeah. So I don't – You played it inside or outside? Outside, yeah. Outside. I mean, two courts is a little bit – you know, I don't know if you love something.

2:58:45Zoom out for me. Build the ultimate mansion for athletics. Are you going all in on one particular athletic endeavor? Are you making sure that there's a pool and a basketball court, a tennis court? What's important to have around the mansion? For me, definitely a basketball court. A basketball court. And then like a gym in a garage with no AC. No AC. And what's in that gym? Are we doing rogue racks? Are we doing? Rogue racks are good. Okay. Yeah, yeah. Free weights? What is it about, what's the archetype that can make a home gym really work? Do you have to be like spiritual? like every time I've just been in a like during COVID I bought a bunch of weights and I was working out of my garage and I just got so small.

2:59:32Hard to be motivated I got jacked, that's a skill issue I'm implying it's a skill issue because I would be going in there and I was at home so I could get an email and I'd be like okay I'm going to run in and respond to this email on my computer because it's easier and if I was taking a more monk like approach I'd probably have been getting even more jacked because there's no distractions right it's just like more focused well during covid i mean we ordered a bunch of weights me and my lifting buddy like right when like right at the tail end because everything sold out like super fast so you still had a buddy that helps had a buddy yeah and um um and we we ended up getting these like dinky probably like 25 pound barbells okay these like shitty weights or whatever um but i would just go and i had a shed at the at the house and I would go in the shed and throw on music.

3:00:21And it was actually the first time that I was lifting where I understood the power of shrieking and screaming while you lift. There's a lot of power. Yeah, yeah, yeah. Okay. The modern man does not shriek and scream. What's in the supplement stack? What's Lindy? What's interesting? What are you staying away from? Are you beta alanine? Are you using creatine, protein supplementation, mass gainers? What do you like? I'm feeling reactionary against the optimizers. I want to just say nothing. I mean, creatine for sure. Protein's fine. But I think you could basically eat raw milk. I do like raw milk.

3:00:59I drink a lot of raw milk. Recently, I went from like 165 to 200 pounds over the course of like three months. And so I was drinking like a gallon of raw milk a day, a tub of Greek yogurt a day. Yeah. The raw eggs, I do think like drinking 12 raw eggs makes you feel amazing. Okay. But that's kind of the - I went through a raw egg. I went through a phase where it would be like five eggs mixed into raw milk, just raw. Were you bodybuilding.com guy back in the day? That was a long, that was a long, that was like, I wasn't even like a phase. I would just, that was just me for a long time. I realized like, wait, I don't have to scramble the eggs.

3:01:36I can just put them in a jar and just drink it. I would, I didn't like the taste. I would kind of almost throw up sometimes, but it was just so effective. I mean, Derek from Our Plays More Days talks about just drinking the full carton of liquid egg whites. He does. Just chugging the carton. Yeah. That was a good era. There was a period where I was watching his videos. Like when I was first getting lifting, watching like – That was great. I remember being like, I can't believe I'm watching an hour and a half videos about something. It's crazy, but it's so meditative. Well, it's like what we were saying at breakfast.

3:02:04It's just amazing to watch anyone who loves what they do. Master of craft. Yeah, exactly. Makes sense. It's great, yeah. You would have appreciated – we gave a talk last year. when we had done like a few episodes of the podcast and then we gave a talk at Hereticon of like the case for like founders getting on like PEDs because you have like when you think about like pro athletes are like spending all their time trying to be healthy, sleeping well, working out and being as strong and as physically fit as possible and they legally cannot take PEDs. Although if you ask them like if you could take PEDs, if it was legal, would you do them?

3:02:41They'd be like, yeah, absolutely. I want to go to the next level. Meanwhile, in the business world, you have people that are not sleeping that well. They can't train that much. They're not that healthy. They have like kind of not that great hormone profiles in a perfect world. They just start sleeping well and training and eating well and getting on track. But for a lot of people, if you're a CEO, busy CEO, you're scaling your company, you don't have like, some of these people like can't prioritize it or they don't for whatever reason. And I genuinely think if they just got on PEDs, they might build the, of course, under doctor's supervision, but they might actually get motivated in order to build some of those healthier habits.

3:03:20But at the very least, have a lot more energy and bring that intensity. What's your take on PEDs? I don't do them because I haven't had kids yet, but they do seem probably good. I don't know. Have you done them? Actually, one of my buddies from Cleveland was having a hard time. I won't say his name. He was having a hard time on the dating apps. And he was like, no one's responding, you know, whatever. And he started blasting gear because he thought it would help him get girls. Did it help him? Yes, immediately. Whoa. And I was like, but he didn't even get jacked. It was literally immediate. Was it a mindset thing?

3:03:57I think it might have been a mindset. Sure, sure, sure. Because he's like, I've solved the problem. Yeah, it was crazy. my problem so i can bring the confidence it worked for him like the next day i i find uh i find dating last year what's that i find creatine genuinely very yeah yeah i take i also find caffeine extremely performance enhancing in the gym like if you can like the celsius phenomenon i think is very real like 200 milligrams of caffeine getting actually very fired up before a workout genuinely helps you stay focused during the lift push your push your Yeah, we want to make the beta alanine for the workplace.

3:04:34Imagine you've got to do some emails. Is that the tingly face? Yeah, you start tingly. You're like, I've got to get through these emails. I mean, I will say that, yeah, for sure. I guess the supplement stack is like drinking enough caffeine to kill a small child. It's really good. Not that I'm going to kill a small child, but that would if they drink it, kill them. These mates are really good. Are you guys sponsored by these guys? We're not, but we're just friends. That's Andrew Huberman's brand. Fuck. Goddamn. Sorry, man. Hey, look, look, look. I love your product. You're like, yeah, I'm talking trash, but you're reminded.

3:05:02They're good. I don't even consider Andrew's philosophy about over-optimization. Nothing you said is wildly disagreeing. Yeah, but yeah, I think it's more about understanding yourself and the various inputs so that you can then make calculated decisions. Yeah. I know how important sleep is. There's plenty of nights. We were traveling on Tuesday, didn't get a lot of sleep, made a call to try to get some more sleep. But ultimately, I think it's just about, my view is like, as much as possible, try to understand how the body works. But at the same time, you need to remember that the real edge is like having like a fiery spirit.

3:05:52Like that is like, that's the thing that you want to cultivate, right? Yeah, there's that. Yeah, I mean, I, yeah, maybe part of my feeling about it is that the one time I tweeted something negative about Andrew Huberman, I got, like, ratioed into oblivion. You know, people calling me gay, and it gets, like, 5 ,000 likes. You know, and I'm just, so, yeah, they came after me hard. But you're probably right, you know. I don't know. But then, but the counterargument to that is, like, that. He makes a delicious yerba mate. Well, yeah, he makes a delicious matina yerba mate. He makes a great product.

3:06:28And actually, in my first novel, there's a long passage about how Yerba Monte is great. Oh, really? Yeah, yeah, yeah. In the novelist. But I guess the counterpoint to the sleep thing that you're saying is that there's that for a while when, like, the Tate brothers were just all over the field. There was that one of Tristan being like, you know, I'll go to sleep at 6 a.m. and wake up at 5.59 a.m. and, you know, one minute before 6 a.m. and I don't need any sleep. And, like, that's the sort of counterargument. That's what you want to do? I mean, it's just. negative one minute. Is that what he's saying?

3:06:56It's an admirable vibe. It's the attitude. Yeah, yeah, exactly. Yeah, yeah, yeah. I always do beef with Brian Johnson about this because I'm just like, don't you think that if you just have great will and a life's purpose, you can live forever, even if you're super unhealthy? And I always cite Charlie Munger and Warren Buffett who eat McDonald's and drink Coca-Cola all day long and live to 99 and 100. They don't look jacked. They don't really do any biohacking, but they seem to be extremely sharp. One thing we probably agree on is, like, I don't think you should let health trackers, like, decide your mood.

3:07:33Totally. Or decide, like, how you're going to approach it. Oh, totally, totally. It's like, I wake up some days, I sleep on an eight sleep, I have, like, a 50. No, the tracker is a competition. I'm still going to the gym. It's not telling me what to do. Well, and it's just, yeah, it's a little friendly competition. It's not tracking me, I'm tracking it. You're tracking it. I'm tracking it. But then there are also studies that are like, you know, the people that live the longest have like, you know, they have like good relationships. Yeah. They have some sort of spiritual life. You know, it's like those.

3:07:58Yeah, that stuff is very unquantified right now in the quantified self movement. And it needs to be more included, I think. Yeah. That's why I don't think you can really like quantify that stuff. People are really uncomfortable with things you can't quantify. Totally, totally, totally. But I think probably a good relationship, some sort of spiritual life, general health, you know, eating the things that like you're. Because also different people, different diets work better for different peoples and stuff like that. Totally. What's your actual routine? Are you in the bro split? What's that? Are you in bro split?

3:08:29Push-pull legs? What are you doing? The manosphere split. Just crushing fentanyl. Not beating the manosphere alligators. Yeah, exactly. Lately, I was doing mad cow, I think it was called. It was just three days a week and it was literally just, it was, um, it was, uh, uh, just power lifting. So it was like squat, bench, deadlift, barbell rows, three days a week, five by five, three days a week. So you do all of those every day, all of those three days or you rotate through them. You, you rotate through them. You're, you're swatting three days a week or benching twice a week. You're doing overhead press too.

3:09:04And you're adding, um, you're adding five pounds a week. Okay. So it was like, and that was when I was doing the, the spirit bowl. And five by fives. Yeah, exactly. I like the five by five. that's good that's a good um so i was doing that recently i i do like i i mean the thing that god it's funny i i the the thing that maybe we're just holding back a little bit of our progress is like we want to work out every day and there's actually like studies that show if you just take more rest days you can put on you can you can just like start but the problem with that program is like i also like lifting every day yeah so it was like a problem like i was like it was actually not the most enjoyable thing i do like as much as i'm like uh opposed to it in theory i do like jeff nippert's programs oh yeah power building program i like him a lot yeah they're they're very good yeah um so the power building one and power building three yeah i think my favorite programs yeah i just like the way he like structures his like lessons and stuff well he's also natural which is like a big deal you know what i mean a lot of people who make these programs aren't natural so it won't work for guys that aren't sure sure sure sure yeah makes sense yeah they're like do 10 ,000 reps of this.

3:10:11The Rich Piana 8-hour arm workout. I'm on enough trend to kill a horse. Yeah, yeah, yeah. That's fantastic. Thanks so much for chatting. Do you have your next book in the works? Yeah, kind of, but I don't want to talk about it. Yeah, yeah. It's going to be a thriller. Can you share the prompt? What's the prompt? No, no, I can't. Don't make mistakes. No, no, I'm not going to do it, but thanks for having me on. Muscle Man is the current one you guys can buy. What about the biggest fish you've ever caught? Dude, I just did. I never. You've never been fishing? No. No, not once. No. Not even a little bit now.

3:10:46My family has. No, dude. I was too punk. I was too much lucky as a skateboarder. You've never shot a fish in a barrel? Nothing. Nope. Nothing. Okay. Well, we'll have to change that soon. Yeah. The outfit looks like it could maybe transport onto a fishing boat. I agree. I agree. Onto a fishing boat and fit in. It's funny. That's funny because I wore this the other day for an event. And my buddy was, I told my friend that I was coming on this show. and he was like, you're actually kind of dressed like, he thought the vest was somehow TBPN. Also, Brandon Content over there has the same vest. He told me about, he told me that he has it.

3:11:14He's looking sharp. Well, thank you so much for coming on the show. Leave us five stars on Apple Podcasts and Spotify. And if you're listening to this and you haven't lifted today, hit the gym. Go for it. Get in the lift. Hit the gym. Visit the Church of Iron. That's right. Have a great weekend, everyone. We love you. Goodbye. See you Monday. All right.