LIVE CLIPS
EpisodeĀ 3-9-2026
Different skill set that comes out in the best customer service agents. How are you thinking about actually driving incremental sales through customer service agents? Yeah, I think one of the interesting things is that a lot of the work in these functions isn't entirely separate. We just needed to draw lines between them when we were running human organizations. Because you can't possibly have one person who's outstanding at service sales, marketing success, et cetera. Yeah. And the market's going to go away with failure. Make the SaaS market maps too complicated. Totally a pain in the ass. But unfortunately that pain in the ass is coming with AI because I just don't believe you're going to have these separate sales agents and marketing agents. You're not going to have these things fighting with each other, having different context and information, different memory in different parts of the customer life cycle, different messengers. Not going to happen. There's going to be vendors that will provide one seamless solution across the customer life cycle. And that's a big thing we're passionate about. And one of the very fascinating things we've seen, we've been building this for quite a number of months now and we're going to launch the first part of it next month. One of the fascinating things we've seen is that when there's an underlying relationship with the agent and then there's actual connection in the moment and then there's clear intent, the context is primed for selling and encouraging the user to do something next. If you're on a website and you get a super annoying pop up that says hey, John or Jordy, in fact, they're not even going to use your name. They're going to say, do you want to learn more about our whatever? And you're like, get out of here. It's like not the right time or place. Whereas if you're actively discussing, it can ask you questions back. And it's so much more effective. And so we've seen just these things. The service agent, when it starts to do sales, for example, and vice versa, they're far more effective. And so the real magic from AI across the board, and of course this agent world is going to be when the agents can do things that humans can't. Yeah. One thing that stands out is like right now you have this spikiness in dialogue as a customer of a business. Right. You might have a question coming in and it creates a moment and then it goes away. And then you get maybe transition to someone else and you're talking to real human. You talk with them a little bit, and you get passed to someone else. And, like, there's something about creating, I think, the customer experience of having this continuous conversation with the organization that you're a customer of that I think will be super powerful because it just means that, you know, a split second away, the entire time I. I was talking,
Fleet, Right. Or is there something I'm missing? Yeah, I'll just add one framework that I think is helpful for people to clarify what that means. Basically, everyone is conflating the model's intelligence with what human are really good at, which is judgment. And I think if you strip those two concepts that you really understand what the future will look like. And so right now the models are becoming really good at everything intelligence related. So think of tasks that are, you know, rules based, where just logic gets you to the goal. Right. Whereas the judgment is everything the humans really spike at, which is more harder to define. Right. It's instinct. People call it taste, call it experience. And those things are things we think the humans will remain really good at for a very long time until intelligence is able to absorb these skills into it. But it's just really hard, right, because we give the models a lot of credit for their intelligence, but they're just trained on set amounts of data. And so we actually think that this hybrid motion of coupling the intelligence from the model with the judgment from the humans is going to help you build the next generation of companies, which we call these autopilot. Right. So to your example in the law firm, you will automate a lot of the intelligence work that's very much codified, which is what the model will do. If you put the employee at the center, someone who's really experienced, who can talk to a jury, who can, you know, mark a pause when at the right time. Those are all the things that you learn only from real world experience. That a top lawyer will be able to experiment, but that the models might, will emit today. Does that make sense? Yeah, I still, I still like, you know, maybe, maybe I'm too AGI pilled, but I, but I just assume that a sufficiently advanced agent would be able to capture and accumulate, you know, the ability to AGI pill. Just Dr. Alex Wisner Gross for 30 minutes and he's like, I'm uploading my brain right now. No, it's.
That's the. All right, well, it's definitely not gonna help us near term. So here's the most interesting thing, which is a pet issue of mine, is we're like, this administration, I should say, does a lot of good stuff on energy. I say that I'm in a fortunate position where I get a lot of people ask for advice and stuff like that. I feel like one of my jobs is to tell people things they're not doing, a lot of things they'll just do right on their own. But Canada is not one of those things. Right now, we are just absolutely sleeping on Canada as an opportunity. So. So Canada has no people and infinite resources and is really friendly. Yeah. So they have these oil sands. You know what these things are, you know, explains. So they're just like these deposits of oil. It's like nature basically committed an oil spill is the way to think of it. Okay, so there's all these oil sands. For years, they were not very useful. They weren't viable. Because it's not just you drill down, black gold comes up. It's like, buried in the sands, need to be replaced. There are different kinds of things. So you can mine it out directly or you can do what's called in situ, which is you heat it up underground. Oh, and it liquefies. In any case, they have just this unbelievable oil seep. They have way more oil deposits than we do, way more oil reserves. But they have on the order of a third of the production. And in part because they have absolutely terrible policies which are their fault. But in part, we have had terrible policies, including getting rid of the Keystone XL pipeline and this kind of thing. But if you just think of Canada, it's just, we should be. So they have water. I mean, again, no people, one tenth the population, infinite raw materials. They got everything. They got timber, they got natural gas, they got oil. And we're talking about Venezuela. They're a friendly country. Right. They got smart people there. They haven't been driven out of the country because they're afraid of getting killed by Chavez or Maduro or whatever. So this is an opportunity to say, hey, Canada, let's work. Let. Let's have a task force. You guys are at a low in terms of. We don't have pipelines, but we can do rail. At least we can do rail. Maybe trucks, but at least rail. Rail transports from Canada are really low right now. So let's have an initiative where maybe you bring a couple hundred thousand. Depending on your expectations, maybe we can add a couple hundred thousand barrels. A day. So that's one that excites me because I want to be like the US Canada, superpower thing is this huge opportunity that we're sleeping on because of the idea of, oh, we should only be doing it in, quote, America. But this is just, by the way, unbelievable. I think they have the best uranium in the world. Yeah, it's maple syrups up there. Yeah. I don't have a Canadian personality, but I love that place. No, strategically. Strategically. And they're very friendly. Yes, they're very friendly. Maybe we should be friendlier. Yes, we should. And we have trade agreements coming up some. The other thing is, we can increase capacity here.
You. Oh, we're already there. I mean we're already in the era of recursive self improvement. All of the Frontier Labs are, are pretty public about it at this point. I'm looking past the singularity, so I spend most of my time on bets for what the post singularity world looks like. So Eon is one of those bets. Another bet. I have a company, Physical Superintelligence, that's trying to solve all of physics with AI and doing an amazing job. I just wrote a book with Peter called Solve Everything, arguing that entire disciplines are going to get steamrolled by superintelligence and that what matters now is what disciplines we aim the superintelligence at. Interesting. What decisions do you make in your personal life, in the way that you live, that are based around your beliefs, around technological progress? There are a few that come to mind. So one is I'm hoping not to die. I. It would be a shame if I died sometime soon. Get to miss the most exciting developments. Same. That's what gets the applause. That's great. Yeah, that's a white pill. White pill, very good. Brian Johnson thanks me in advance. But I think also, I mean there are so many other angles. Another is not trying to bet against the collective intelligence of the market. I've had a number of friends who thought they could outsmart all of the AI algo traders and day trade and when they inevitably fail, they attribute it to bad luck or something other than just fundamentally betting against progress and betting against AI intelligence. So trying not to make the mistake of betting against collective intelligence of civilization and AI capabilities. I'm trying to make bets, assuming that the AI capabilities keep increasing to the point where they're ontologically shocking and trying not to duplicate effort of Frontier Labs. And I think that's perhaps something I don't see enough of in the venture community. So many new startups being formed now just aren't being ambitious enough. I have startups that newly formed or recently formed that are literally trying to grow new islands and new coastlines with AI that would have been unthinkable years ago. I have. What is that's with physical AI or they're just dreaming them up? It's with AI reaching into the physical world to steer ocean currents and grow new islands and new coastlines. That would not have been possible a few years ago. That's now starting to become possible. I have a startup that's working on solving interspecies communication, starting with dogs. That would not have been possible a few years ago. So I spend a lot of my time thinking. Thinking about what the post singularity state of the world looks like and how to smooth that out and bring it here sooner. Yeah. What does the post singularity world look like?
$2,000 in compute. Now the same $200 a monthly, monthly plan can consume 5,000 in compute. It does feel like there's a capital war. I'm very interested to dig into the history of capital wars in winner take all markets, but also in oligopoly markets. You should write the book on capital wars. Capital wars are fun. Coogan. I lived through it. During the Uber and Lyft era. There was a flippening for a little bit where Lyft was, was at the top of the charts. And the friendly, the friendly rideshare company, it was very friendly. The cars would pull up with a pink mustache on them and it had a very, very fun, whimsical branding. And you would sit in the front seat of the car because you were. I can see Claude. Did you notice a pink mustache is a real thing? Yeah. So you're supposed to do so when Lyft. So when Lyft launched. People see Lyft and Uber as like commodities at the time, but Lyft was like revolutionary. And people were super black belt on Uber because Uber at the time was only black cars. So they didn't allow anyone to just hop on the network. Like it was not a three sided marketplace or whatever, two sided marketplace. Like you had to have a, like a livery TCV number, I forget what it's called, but like you had to be registered as a limo driver and you had an Escalade and normally someone would call you and say, hey, there's a wedding and we want you to come shuttle people around at this wedding and they would pay you for that. And Uber went to those folks and they said, hey, you are red registered to legally drive this car. As a professional driver, we want to book you through our app. And so the black car became Uber's main product, their only product. And then Lyft figured out through some interesting, you know, strategy how to allow people to just show up with a Prius or a Toyota Camry or Corolla or whatever. And so drivers on launch day, I remember like the launch week in San Francisco, would just show up in just some random car. But Lyft gave them all pink mustaches, huge pink mustaches. The team can pull up a picture of the huge pink mustaches on the front of the Lyft cars that were very whimsical. And it was remarkably cheap because Uber had been somewhat subsidizing the black cars. They hadn't been taking a huge margin, they were losing money. But the Lyft cars were way cheaper because someone was just showing up with like the, their Toyota Corolla or whatever with their pink mustache on it. And so you would hop in the front seat because this driver was your friend. There you go. The pink mustache on top of the. What is that, a CRV or something? And so you would hop in the front seat, they would ask you to pound the driver, give him the knuckles, give him knucks, give him knucks. This is the thing. No way. I'm almost sure. So you'd be like, what's up? And then you'd get in the front seat and they would drive you. Of course, if you had a group, you'd be in the backseat, but then everyone had to give knucks or just. I think the knucks might have been optional but encouraged, recommended because it was like, hey, we're buddies and we're, and we're fighting the war on the Borg. We're going up against the border. But you were encouraged to get in the front seat. Front seat, front seat. In the lift. In the lift. When you're taking a lift, your friendly Lyft driver shows up, you sit in the front seat, you give him nugs. This is the thing. Pink mustache. Chungus, big Chungus sends his regards. All right, so you got to hire a car, but you have to be absolute boys. You do. But, but truth be told, like a lot of the early Lyft drivers were like super cool tech forward San Francisco people. And so like you'd have a lot to talk about because people were, it wasn't professionalized at all yet. It was very much like two sided marketplace. You'd make a lot of money as a driver. There was no margin compression, yet it was free money everywhere for everyone. And it was a capital war. And Lyft raised a ton of money and came into the market with a disruptively cheap product and wound up delivering a really fun experience that was very, very cheap. And then Uber sort of fast followed and launched Uber, what's it called? The normal one. Not black. Uber X Ray. Yes, Uber X. And Uber X was bring your own car. And then, and then the rest of the game played out and it became a capital fight. And for a while like the meta was oh, you're, you're, you're, you need a lit, you need a, you need a ride. You open Lyft and you check the price and then you open Uber and you check the price and people would check both prices constantly. And then, and then over time, like the liquidity and power law stuff came into John do you remember Uber helicopter? Yes, that. Was it a real thing with Blade? I think it was a partnership with Blade, which was the Uber for helicopters. And it was a thing, but in. In like, you know, for a few weeks or something. Did you ever do it? No, no, because Uber. Uber's had a whole bunch of different, like, stunty products like Uber, you know, Uber yachts or Uber. They did Uber ice cream, where you could get ice cream delivered. That was like the precursor to Uber Eats. They ice cream delivery truck where every Uber driver checked in with the field office, got a cooler in the back with a bunch of dry ice and a bunch of ice cream, and you could press a button and for like $2, they would bring you an ice cream. They even did, I think, Uber kittens. And you could ask, yeah, Uber dogs or something. I remember that. Yeah, it was very funny. Anyway, lots of stunts, lots of ways to get attention, go viral. And that was sort of the. The launch strategy. They would do that for whenever they launch new markets, they would come in and they'd be like, we need to go viral on the local news. We need to let the town that know that we've arrived. And so lots of back and forth in the Capital wars of 2,023, 2,024. It was fun times. I lived, I fought through them.
Against chaos in automatable software. All of that. Yes, I will say this with love to all of my LPs. I think founders are typically like leading indicators in categories like they they are at the cutting edge. Like they are seeing around corners often like great founders will see something before VCs. Like VCs are lagging indicators. Like once something is hypey it's like usually too late to find a good early stage investment. LPs are lagging, lagging indicators that once something is very mainstream and there starts to be some signs of returns then they're like we should pay attention to this category. And so I would say a thing that was more surprising partly because I live in my own echo chamber and constantly hear about all of these great companies building in defense tech, hard tech and industrial tech. LPs weren't as familiar as I expected. So there was a level of education. I think there's still some wariness of. There haven't been a ton of exits yet. I mean I think that might change if SpaceX IPOs will be one of the biggest IPOs ever in history. But I would say like the appetite for it grew over time as I was fundraising on SpaceX.
Table and then the question is you need to be able to produce it and then you need to be able to transport it. So you have estimates around maybe you could get 1 to 2 million barrels a day. So that's one of them. In terms of emergency, if you look around the world, you don't have a lot of immediate spare capacity besides that. So we have what's called emergency capacity or Strategic Petroleum Reserve. It's Jordy mentioned that we didn't fill up our Strategic Petroleum Reserve. I think this is unequivocally a mistake. I mean, if you look at the whole thing of the Strategic Petroleum Reserve is you want it for when you need it, and the best possible time to fill it up is when prices are low. And is that physically a place with a bunch of barrels that we just. There's a bunch of different knocks of oil? Well, there's multiple different places, but you can think of it. But we basically cash the physical oil in America or somewhere where we own it or can access it. Yes, exactly. And it's sitting there going unused until we're ready. Yeah. Now there's issues of Biden. Biden definitely misused it. So they used it to basically have lower gasoline prices during midterm elections, even though parenthetically their entire policy goal was get rid of fossil fuels, which means you want the price to go up so people can't afford it. It was a very cynical kind of move. They did it in a way that degraded the facilities. So let's say we're at about 4 million barrels a day. We can get out of that thing. We have 400 million in there. We should have over 700 million. We could have easily filled it up. I mean, it was a perfect time. When you're talking about $50 barrel oil, $60 a barrel oil now. And can I do the basic arithmetic of 4 million a day, 400 million in the reserve, 100 days of oil. But that's not actually if we're talking about like relieving price pressure, you could. You could trickle it out over a year and have 1/20 or 1/4 of the effect. But just think about $100 million a day. 100 million barrels a day is roughly the global market. Okay, okay. So. So 4 million is just. That's its maximum throughput. Got it. It's kind of an analogy to batteries. We're talking about batteries. So there's a certain amount, like with a battery, if you hear there's a four gig, there's a one gigawatt battery installation. That means that's the. It usually means there's one gigawatt for four hours. Sure. But if you wanted to do half a gigawatt, then you could do eight hours. Right. It's the same deal with this. But keep in mind, this is not 100 days of US oil demand. And by the way, for the refinery reasons, we can't just supply it all with our SPR. It's just the maximum output is. But 4 million barrels a day is something. We're talking, that's one fifth of what's flowing through the Strait of Hormuz. So yeah, it's a thing, but yeah, you can think of, yeah, maybe we'd be willing to do 1 to 2 million barrels a day from ours. Now, interestingly, the International Energy Agency, we are part of the International Energy Agency reserve program. So we have about 400, but they have about 1.4. 400 million. They have 1.4 billion. So we have allies around the world that can also release this oil. So maybe we could get 1 to 2 million from them now, at least last I checked the news, they hadn't agreed to do this because they don't think it's an emergency. Sure. But that's what it's there for. Yeah. So notice the pattern is at what point would they think it's an emergency? I don't know. I mean, and we, you know, we can put pressure on them. I think some of these determinations have to be made in conjunction with the military determinations. You can't make them in isolation. Because if you think about what's the military objective? How close are we to that? Or how close are we at least to opening this up via high security and financially backed convoys? Then you can think about the timetable here. If you think about, well, the Strait of Hormuz is going to be closed for five years. Let's just get used to being poor, Black Bill. No, I mean, this is.
Sure. To get back to that one. But the main thing is just rough numbers. Straight up. Hormuz, which Iran has control over, is 20% of the world's oil production is flowing through that every day. And so whether you're looking backward or looking forward, there is no replacement whatsoever for opening that straight and keeping it open. And one of the things we've heard from different people is, oh, we have a lot of options on the table. There's a lot of things that we can do. There are definite things that you can do, but none of them is 20 million. Right. So it's just. I mean, that's. That's more than U.S. oil production just flowing through that one place. So, yeah, we can talk about the other things, but I think the first thing people need to recognize is if you do not get that thing open, then you have dramatically higher oil prices. And then the other thing to get is oil prices really matters. Dumb question. Can't you just go around? I mean, you can try to some extent, but there's a reason why 20 million a day are being routed through, because it actually comes out of there as well, Right? Yeah, there's just. I mean, all of these things are very, very optimized. One of the thing about just understanding the energy industry is like, things are just very optimized in terms of where the infrastructure is. You're doing all of these things. So it's. There are other. And we'll talk about this. I mean, there are certain pipelines, you can pipe more oil through there, alternate routes, but we're talking about in the millions a day. Maybe most of our other options are sort of in the 1 to 2 million barrels a day. And just so people know, barrels, 42 gallons. So you're talking about almost 100 million gallons a day for this stuff. But, yeah, so you don't have a lot of great options. And I think that's very important for. I'm going to give some options. But the number one option is keep that thing open. And there are basically two ways to do that. One is you just win the war, if you want to call it that, a lot more quickly. Like, if you get some form of surrender and you have friendly people controlling it, then guess what, it can open that. That's one thing. So I'm not an expert on how to do that, but schematically doing that is very effective. The other thing, which is a little bit. But game. Game theory for Iran is they actually.
Systems to push that part a little bit to the left so that the AIs don't have the Dyson swarm all to themselves. Interesting. Architecturally, what does the neural network look like that you've built for the fly? Is it similar to a large language model like transformer based architecture? Or are you a beneficiary? Is all you need. Totally unlike the transformer, totally unlike the transformation, totally unlike the transformer. So transformer architecture, if you subtract off the encoder and decoder layers, looks comes in many variants, but the most vanilla variant looks like alternating linear and attention layers. So totally unlike the transformer architecture. At least if you look at it, it looks a little bit more like a graph neural network. But really that's such an AI way of framing it. It looks like a leaky integrate and fire LIF model that we've had for decades from the neuroscience world. So it's just a graph of nodes that have leaky intranet and fire dynamics and they're firing at each other. So what does that tell you about the nature of synthetic human intelligence? Like, do you think that there is a path to human level AGI that does not involve the transformer and instead falls along your path and your architecture? Oh, sure. Well, first of all, I would argue we've had AGI for at least six years now. Five to six years at the very latest. Since summer of 2020 when OpenAI published their language models or few shot learners paper. Yeah, GPT3. I would argue that was an AGI. So I think there are many ways. My bet is there are many ways to achieve generality of intelligence. I published a paper a number of years ago arguing that not only is intelligent behavior a general process, I went further and argued that it is a general physical process that you can even formulate intelligent behavior in pure thermodynamic terms. So my bet is intelligence is this very, very general effect and lots of ways to implement it. Interesting. I mean, you've mentioned AGI six years ago singularity here and you're smoothing it. Do you have other sort of binary benchmarks that you're looking towards?
Yeah. Take me through your current thesis on how software is changing in the age of AI. In the age of maybe some sort of takeoff. AGI is here. According to Sequoia Capital, amongst many other people, the labs are recursively self improving. And it seems like Sam Altman made the claim that if you want to build a startup, you shouldn't build a company that is in the direct path. While more and more products seem to be in the path of AI and AGI, how are you thinking about the changing dynamic with software as we know it? Yeah, that's a really important question. And I think we've been listening to Sam a lot as shareholders and recently just talking to our founders. A lot of them are wondering if they're just an iteration away from the models, replacing what they're doing. Yeah. And it puts us in the position of thinking what is coming next, what is going to be defensible. And for us, we've been making that prediction last week that the next trillion dollar company will be a software business that masquerades as a services firm. And we shared that thesis with everyone online. The reality is if you sell tools today, you're really in the line of sight for the models and you're effectively competing with the next generation that they're going to launch. Whereas if you sell the work, you're actually benefiting from what the models are doing and all the billions of dollars that are going towards AI. So part of that prediction was saying for every dollar that's spent on software, $6 are spent on services. Until now, we could only really go after the $1 because we were building tools. Now that we can actually deliver outcomes, companies should think about capturing those $6 instead. So.
How different labs are pricing tokens and plans potentially to take market share from each other. There's a little bit of a capital fight playing out. How are you communicating more specifically? Competing with application layer? Yes. And so my question is, how are you talking to business leaders CTOs about how they should think about forecasting token budgets and spend and compute budgets if they're used to, okay, I have licenses and licenses maybe go up in price and my head count goes up and I'm used to that. But all of a sudden it's like, well, everyone's paying $200 a month, but it might be $5,000 a month in a couple months. But also the models are getting cheaper and the inference is getting cheaper and the capabilities are going up and it's very hard to predict. What are you talking to? What are you telling to business leaders and CTOs? Yeah. So the first thing is you have to embrace this technology, otherwise you're going to be left behind right now because the productivity gains are just so significant. And if you go talk to a coder and you told them, hey, I'm going to take away like GitHub, Copilot and the agenta coding capabilities, they'd be like, I refuse to work in this environment. Like, it's just, it's just inhumane almost. So I think the same type of thing is going to happen for all information work, all office work. So nine months from now, I think if you went to somebody and said, we're going to take away your agentic tools like Copilot, Cowork, they'd be like, no way, I'm not going to go back to the old way of working. So I think the first is there's a degree of inevitability because the benefit is so large and there's such strong pull from the end users that said everybody has budgets. So what you are going to see and what you're already starting to see is a bunch of controls where IT teams can start to manage exactly how many tokens or how much dollars each user can spend. And they're going to allocate that in a way based on where they think they're getting return and diminishing return. And it'll probably be like expense report approval hierarchy type thing. So you'll see all of that show up. But part of just doing work now is you need a token budget. That's the future of how we think about how you budget for a team or a department. Do you have any exciting case studies on the.
Fahrenheit on Thursday and Friday in March, which is why not good. We're going to do the weather segment. We have a new segment for you today. We're doing the weather on tvpn. We have our very own Ben. We have been. Hello, guys, how are you? We're doing great. Tell us about the weather. What's happening? Well, I want to start off by saying, as you can see, the weather today for the low. Low today is going to be 75 degrees Fahrenheit up there, high of 100 degrees Fahrenheit down there. But there's something I actually wanted to point out that I saw and I thought that was quite interesting. As you can see up here, there's a localized low pressure area up there and a localized high pressure area right down there, if you can see that. Right. Does that mean rain? The issue. That's not normally an issue and not a cause for concern. And it's not very common for this time of the year. However, today, later in the afternoon, these two areas are going to collide and they're going to hit each other. Really? And what that's going to cause is a barometric pressure inversion. Okay, it might sound a little bit scary, but I guarantee there's no cause for concern. All that means is that hot air rushing in from the west is gonna collide with that cold air rushing in from the east and it's gonna cause a bunch of turbulence in the sky, moving all the airwaves around in oscillations in the sky. However, I wanna add one more point. A byproduct of this effect is that all that humidity that dropped after that hot air moved to the bottom is gonna raise up, cause the water cycle, you know, evaporation and stuff. It's gonna raise up into the sky, into those clouds. It's gonna cause big clouds in the sky and eventually all that water is gonna fall down onto the ground. We're gonna have big rain later in the afternoon. Wait, it's actually. Yeah, it might sound crazy, but I just want it for all you guys at home. I'd definitely try to step outside with a jacket today. Maybe a hoodie just in case the rain comes. Whoa. Don't try to be a big shot. I'm fact checking you right now. And the weather app says it's going to be sunny all week. Is this just complete fake news? My team. Those are the numbers my team gave me. This is complete fake news. No, no, no, no. This is the fakest news I've ever heard. All that Corn. Literally. The transatlantic current. John, John, I don't trust your app. I trust the weather. Apple says it's not going to rain the entire month. There's zero chance of rain. Did you look at the transatlantic current? No precipitation. Zero inches today. That's tomorrow. Zero inches on Wednesday. John, you're really going to. You're really going to trust. You're really going to trust an application. Yeah, it was probably vibe coded yesterday. Over Ben, who's doing the weather. How did he get here? What happened here? No, if you look right there, you can see a localized high pressure area. That's a localized area. This is an over eager weatherman who's just looking for drama in the most boring weather market in America, which is Los Angeles. I think this is the most important story in the world. This is ridiculous. You guys can look at the jet streams. They're coming in from the west. I don't want to hear any more mumbo jumbo about jet streams. Get out of here. You're done. Great work, Ben. Thank you, Ben. The fake news weather will return. Maybe never. Who knows? Let me tell you about Restream 1 livestream. 30 plus destinations. If you want to multi stream, go to restream.com and let me also tell you about Gemini 3.1 Pro. With a more capable baseline, it's great for super complex tasks like visualizing difficult concepts, synthesizing data into a CPU, or bringing creative projects to. This is the future of the weather. You have a weatherman who gets into a live altercation with one of the other hosts who's just constantly fact checking it's on site. Next time, Ben, it's on site. If you pull up with some fake news on TVPN's weather report, it's on site. Ben, you crushed it. Good job.
like a time tested time honored economic philosophy that is deliberately not revolutionary and i thought that it was a great book for that reason and that's why i recommend stubborn attachments to so many people and so widely but if you want to crack open the real hard stuff pick up a copy of the wealth of nations why is one of those copies so much smaller is the font size just smaller yeah it's super small it's also two columns oh two columns wow yeah if you can get through that you're truly elite in the global economy what does the timeline have to say about the masterwork that is turning two hundred fifty it's now needed more than ever says the washington post on march ninth seventeen seventy six four months before the american colonies broke with britain over the issue of taxation a little known scottish thinker published a long dense book with an unpromising title shots fired an inquiry into the nature and causes of the wealth of nations two hundred and fifty years later adam smith is by any objective measure easily the most widely cited and widely quoted economist who ever lived astonishingly his work still frames the central questions we face not just about free markets trade and capitalism but but about the nature of human society itself and even what it is to be human at all smith was born in seventeen twenty three in kirkcaldy scotland educated at the universities of glasgow and oxford he initially made his name not as an economist but as a moral philosopher his first published book the theory of moral sentiments offered a radical theory of how we form moral judgments radical because it derived from the creation of moral values not from scripture or divine grace but from human sympathy and mutual regard the wealth of nations as his second major work came to be known was an extension of that product the book is not as sometimes believed a hymn to greed a pain to market fundamentalism or red in the tooth claw capitalism it was an attempt to understand how a commercial society could generate prosperity without collapsing into corruption this two hundred fiftieth anniversary is not a moment for hagiography it is an opportunity to recover a way of thinking that is directly relevant indeed urgent to the economic social and political challenges we face begin with trade in his own time smith's great target was mercantilism or what he called the mercantile system the idea that wealth consists of hoarded bullion and that trade is a zero sum contest governments granted monopolies imposed tariffs and manipulated commerce in the purported pursuit of national strength producers were widely protected consumers often ignored altogether against this smith one of these days we should have tyler just do a speed reading of the wealth of nations live on the show as long as you speedrun yeah wealth of nations speedrun any percent glitchless you need to do glitchless though glitchless for sure a glitchless run is the only way we can do it no glitching no ai summaries i think that would take you probably five days probably five days it's so dense have you seen how fast you can rip a rubik's cube this edition is four hundred eighty eight pages right oh yeah i guess you don't have to read all the footnotes and stuff you can kind of skip through it but it is a thick book but the way it's written is just really not fun to read it's like it reads like it's two hundred fifty years old it's not like a page turner well we need to talk about the wealth but first vibe co we're dtc brands b two b startups and ai companies advertise on streaming tv pick channels target audiences and measure sales just like on meta and let's also tell you about fin ai the number one ai agent for customer service if you want want ai to handle your customer support go to fin ai when you talk about the wealth of sundar pitch i pitch ai sundar new pay deal worth up to six hundred ninety two million is this like ten times what tim cook's making no it must be overtime right i mean cook's making around seventy a year seventy a year combined so if he works for ten years he makes what sundar makes in three there's going to be a conversation it is i mean but this is what we've been we've been advocating for this so this is good yeah no no we're in support but google has increased sundar's potential pay to six hundred ninety two million over the next three years you know that tim cook dropped this in the apple board members group chat as soon as it hit he was just like dude this is a cool article you should read this drops it in just drops it yeah exactly the bulk of his package comes in performance units with a target value of one hundred twenty six million split evenly into two branches the psus are valued by the parent company blah blah blah it could pay out as much as twice the target or a quarter million if it outperforms significantly or nothing if it lags behind gotta beat the s and p one hundred pichai will receive waymo stock with a target stock in waymo with a target value of one hundred thirty million and forty five million in wing aviation that's their their drone delivery platform platform again both can pay out up to two hundred percent of the target they've got a lot of incentive pay like if he delivers and the company does well he should be richly rewarded behrens was writing about the death of the mag seven mag seven's over up and down wall street the mag seven era ends the barron's is calling it they said stick a fork in it turn out the lights hasta la vista say it any way you'd like the simple truth is the mag seven trade is over finito i love barron's writing dead the collective stock market outperformance of those seven tech icons alphabet amazon dot com comma apple meta platforms nvidia microsoft and tesla is now a thing of the past the group may still do okay and some of the individual stocks may even kill it but the slam dunk set it and forget it run circles around the market era of the mag seven is gone with the wind recall that b of a securities analyst michael harnett coined the term mag seven in twenty twenty three pretty recently referring to the nineteen seventies john sturges western gunslinger flick an adaptation of akira kurosawa's nineteen fifty four film seven samurai have you seen seven samurai no have you seen the magnificent seven the stocks no the movie the movie the western no it's down the seven samurai in what world production team has anyone seen either of those movies i think we watched them in school okay okay you're saying you didn't watch them you studied them haven't seen them in a while which ones have you seen tyler seven samurai seven samurai it's a great one it's a very it's a very good movie anyway barron's is bearish on the mag seven but sounds like sounds like it does sound like hope i think the mag seven will do it's like the large it's over for the largest and most profitable companies in history that stand to benefit yeah the some ways the actual argument was the growth to value narrative the loss of cash flow as they increasingly invest in capex the financials will look very different there'll be a margin compression that type of of thing it's not an unreasonable take but it is just funny the way it's written anyway sundar pichai took over as ceo in august of twenty fifteen he's going on eleven years in the seat google's market cap has increased almost sevenfold from half a trillion to three point six trillion briefly topping four trillion in january this surge has made the indian born fifty three year old former mckinsey consultant a billionaire he joined in two thousand four and made his name developing the chrome browser and leading the android division he had been criticized for being too slow to adopt ai at the search giant allowing openai to release the first hit product chatgpt in late twenty twenty two but has since bounced back releasing cutting edge ai models and integrating the technology into its dominant search engine yes he's done very very well on that front pichai has also navigated a duo of antitrust cases brought against google's search and app store businesses avoiding the worst case scen of forced breakup a third lawsuit is pending against the advertising network tyler do you think that deepmind counts as rsi capable at this point recursive self improvement so there was a take on the timeline where basically someone was making the argument that xai is cooked because while anthropic is using claude to build claude and openai is using codecs to build codecs so both of those leading private ai labs are recursively self improving xai was using claude code to build grok and so they don't have the same flywheel of the tools that build the tools build the tools yeah i think they have access to cloud code but then they still have access to codecs okay so you can it's kind of like do you count that as rsi because it's not like really internal yeah but is gemini that's the question yeah yeah i'm sure people at gemini are using gemini i mean there is a gemini cli people don't talk about it as much yeah yeah i mean the model is very good right yeah yeah i assume it is helpful it would be interesting i feel like we don't get as many little data points from the gemini and deepmind team because it's a public company and they need to maybe break out financials at some point yeah i think like vague posting is probably more heavily looked at exactly exactly because you would yeah because it's totally possible that they're doing the same thing i mean demis has been extremely just very clear about the path this goes and he's very agi appealed he understands the exponentials and whatnot so he certainly messaged it at a high level without actually saying definitively yes like we are at a point where we're not writing any code by hand that type of phrasing that you see in the rsi crowd anyway pichai last got a stock award in december of twenty twenty two worth two hundred eighteen million dollars which was structured in the same way his earnings are topped up by his personal security costs which rose to eight point three million in twenty twenty four earlier in the week he sold thirty two thousand five hundred class c shares at an average price of three hundred three dollars worth roughly ten million dollars the bloomberg billionaires index estimates that he has sold about six hundred fifty million dollars in stock since becoming ceo they still own along with his wife he owns one point six seven million shares of google worth half a billion at the latest stock price and google's founders sergey brin and larry page still control the company through their ownership of super voting class b stock which gives them fifty six percent of decision making power interesting i think you got to give miami paper hands he sold six hundred fifty million of stock no no no he sold that but he's oh and he only owns half a billion so yeah he sold more than hopefully he put it all in video shy but he's getting topped up and he's diamond handings the new he's diamond handing i know i'm just saying i'm just saying looking at the stock charts since twenty fifteen when he became ceo would have done pretty well just to not do anything and trust his own process hey hey maybe he has amazing car collection or something you know there's a lot of did it rotate into metal mate you can't drive a google class class a share you can't sleep on a bed of class c shares jordy you got to get liquid no you got to give miami tech its due because everyone's saying oh miami tech they don't have any startups there's no funding there miami tech isn't panning out well google now counts as a miami tech company because the google founders bought houses there you also got to count meta as a miami tech company because mark zuckerberg has a house there so you put meta and google that's like six trillion dollars in market cap for tech companies all in miami i count it i count it anyway moving on owner of four loko is exploring a sale of the storied alcohol brand sources say wow let's dig into this but first let me tell you about plaid plaid powers the apps you use to spend save borrow and invest securely connecting bank accounts to move money fight fraud and improve lending now with ai and let me also tell you about cognition they're the makers of devon the ai software engineer crush your backlog with your personal ai engineering team so the parent company of four loko the canned alcoholic beverage that became a college campus sensation in the late two thousand before being reformulated under regulatory pressure is exploring i cannot believe they nerfed for loca it's too powerful it was powerful like you can just imagine the trajectory of the united states states if it hadn't been nerfed and then straight downward really yes it was so bad it was so i thought people were having like heart attacks yes it was terrible so four loko was a twenty four ounce can so twice the size of a normal alcoholic beverage trey the birthday boy says it was amazing it was amazing okay now we have buzz balls that's true so four loko the original formulation was twice as big as a normal can of bud light or miller lite or coors light your normal beer something that you would grab so it was two drinks that way and then instead of it being three point two percent alcohol or four point five percent alcohol it was like ten percent alcohol it was the strength of like wine almost and so a single four loko was like four or five beers in one can and then they also added like two hundred milligrams of caffeine so you would become incredibly intoxicated and inebriated but then also incredibly stimulated from all the caffeine and that spelled doom for many people people would be very high functioning but completely inebriated and discombobulated and so they would get into all sorts of trouble whether or not it was behind a motor vehicle or a discombobulator it was the discombobulator it was also wildly illegal from an fd perspective you can't mix alcohol and caffeine in a single product that's just a rule that they chose not to follow and they figured out how to make it and then the fda eventually gave them a warning letter and they pulled the caffeine out and that of course killed the viral sensation that was four loko but apparently the company continued to sell products that were compliant with the current fda regulation nerf that's nerfed the nerf for loco continued to somehow they're looking to and potentially will capture somewhere around four hundred million on the sale the brand value so people it's hilarious that they're working with jp morgan on this that's wild so fusion products is the owner of the brand they're working with jp morgan on the sale process the potential sale underscores how ready to drink beverages have emerged as a growth category in an otherwise sluggish alcohol market us beer wine and spirits sales declined declined in twenty twenty five we've talked to a number of guests about the declining alcohol sales as folks move over to more functional beverages here's what the cmo had to say for over a decade our sales have been the leading flavored malted beverage have been leading the flavored malted beverage market by embracing bold innovation unconventional marketing and a risk taking attitude that delivers results year after year i guess they kept building they also own the parent company also owns pirate water pirate water basico tequila and just earthquake earthquake earthquake seems like something you definitely want from the makers of four loko earthquake pirate water is a wild name for a brand too i mean it's a distinctive can it looks remarkable it jumps off the shelves when the four loko nerf went into effect and they pulled them from the shelves because before they reformulated without the caffeine they had to pull them from the shelves entirely they were a hot commodity and folks would go around stockpiling them before they were banned and they were one of the most guarded resources in all of college campuses wow earthquake is just a can that's just ten percent alcohol wow do not recommend very risky never quake anyway in twenty ten the fda sent warning letters to caffeinated alcohol beverage makers including four loko saying that the caffeine in their products constitute an unsafe food additive so the fda has specific categories for regulation around around what is an alcoholic beverage what is a caffeinated beverage and they never the two shall mix so they don't let you put these two things together of course vodka red bull was the precursor to this where folks would order a red bull and then they would put vodka in it i think but at a bar which is made at the time of sale and so that was legal i believe and this might be fake news but i remember hearing a story that at one point red bull was sued because a gentleman had purchased a series of vodka red bulls become incredibly inebriated gotten behind the wheel of the car of a car got into some sort of accident and wound up in front of a judge and the argument was that he wanted red bull to be held liable for his actions while inebriated behind the motor vehicle and his argument was that it was only because of the product red bull had been mixed with the vodka that he made the decision that he was alert enough stimulated enough from the caffeine to get behind the car wheel in the first place and this might be apocryphal but the story goes that red bull made the argument that the caffeine actually made him more capable to drive not less capable to drive and actually increase the safety of the of the driver and the vehicle and so they were held not responsible but i don't know that's just like an old wives tale in cpg i don't know if it's true anyway let's move on to phantom cash find your wallet without exchanges or middlemen and spend with the phantom card and let me also tell you about label box rl environments voice robotics evals and expert human data label box is the data factory behind the world's leading ai teams so we're all preparing for the singularity what is this doing dishes just like oh if you're doing dishes you're not worried about dylan patel's viral post saying that being in sf is like being in wuhan right before the pandemic something is happening it's going to hit everywhere but so few people know it and that is the question of are we in a software only singularity where the dishwasher is life is normal and every time i go to the COVID analogy i start a clock because i remember tracking covid in january and i remember the viral post biology posted in february going viral and then in march the nba shut down and in april everyone was locked down and so every time i see one of these it's like covid i'm like okay let's see what it's like in eight weeks like eight weeks is not that long in eight weeks if i go outside and things look radically different and you talk to random people and it's wildly different then we're like okay yes that analogy was correct but if it's more like in eight weeks we're like oh yeah technology's getting more exciting and tech companies are changing and the market's moving but the dishwasher is unaffected the pizza place on the corner is unaffected that's always how i just think people underrated how crazy a global plague was like it's literally like a once in a hundred like the most insane thing ever it's the craziest change to literally everything from the sports that you watch you couldn't go to the movie theater you couldn't go to dinner like everything was changed for like a full year it was like complete upheaval in our society and might happen but eight weeks eight weeks is a little fast for me and if you live in la probably still be able to go to the beach i think so so that's but maybe not maybe everything's i did have a funny moment yesterday i was telling you but i'll share it again i was talking with my my car detailer who's become a dear friend and he was telling me that the people in my area cancel their appointments at a much higher rate than the people in the rest of la and he was talking about running analytics imessage and i was like how are you how do you do how do you do that and he's like he's fully running open claw so he's now running using openclaw to run his like car detailing business he's like yeah just just always running in the background it's it's some to him he was like it's some combination of having like an ea yep or a ceo like helping me manage the calendar helping with comms helping me make smarter decisions on routing and scheduling and all these things so when you book a detailing meaning that this gentleman is coming to your house and cleaning your car you send an imessage to him correct and that is what is so interesting about openclaw is that there were probably one thousand companies that were like we're vertical saas for auto detailing you got to do everything over email and they just got eaten alive because people use imessage and imessage has been this closed ecosystem that private companies could never access before but it makes so much sense for him to say hey run through all my imessages and if i'm talking to my friends about the f one race just discard that but if i'm talking to a client about potentially booking a detailing put that in a spreadsheet organize that and then run some analytics and tell me what the patterns are and that's just something that should have been saas for a decade but apple has a walled garden that has now been smashed down by openclaw and i think it's really cool and i think that that's what we're seeing the first like use cases for openclaw it's like things that could not happen for business reasons for logical reasons sometimes for legal reasons if you're talking about downloading illegal films like the napster analogy but a lot of times it's just it's just apple didn't want to do a deal to integrate imessage into salesforce the way to break down the walled garden was just to have a digital guy basically yeah who just who just combs over all this so is that a new type of software that that should have existed that now is unlocked is it disruptive like it's not like he it's not like he stopped using his previous crm probably didn't have anything so now he has something yeah he he's pretty tech forward he's tried every single saas tool that there is but like at the end of the day having like a digital guy that has access to all of your tools is just better experience and so i want to know what his bill is he said he spent a decent amount i want to know what it is this month and then i want to know what it is in like a year because sometimes the maintenance of these systems can be really really big and then there's a question about are we going to see per token deflation or are we going to see prices are prices being artificially held low because of competitive dynamics so in a year will he be spending more or less will he still be getting the value these are all interesting questions but i'm just excited that a digital product now exists that could not or didn't exist before and that feels incredibly positive and cool so very exciting anyway we have our first guest of the show in the restream waiting room let me tell you about console dot com console builds ai agents that seventy percent of it hr and finance support giving employees instant resolution for access requests and password resets and without further ado doctor alex wisdom gross welcome to the show how are you doing what's happening doing really well thank you john thank you jordi thank you so much for taking the time to come chat with us congratulations on all the virality for those who haven't followed along so far can you introduce yourself and the project broadly and then i have a ton of questions but i'll let you kick it off with an intro absolutely so quick background on myself originally from new york undergrad mit studied physics electrical engineering computer science math phd physics harvard was there anything you didn't study you studied everything didn't study the humanities okay considered going to cambridge to study chemistry and biology for a while to start at it was told physics is a game for the young okay so stayed in the us did my phd in physics focusing on ai and nanotech i've started invested in advised managed dozens of companies at this point i've had a number of exits and at this point most of my time is focused on smoothing out the singularity as you guys were just discussing so most recently as you were gesturing a company that i helped found eon systems public benefit corporation just announced what we've been characterizing as the first multi behavior upload if you will more on that in a second of a fruit fly which we think is a major step forward for the field and quite frankly in an era when i do another podcast the moonshots podcast with my friend peter i talk all the time about tiling the earth with compute right now all of this ai infra buildout that we're doing to the tunes of trillions of dollars of capex all of this is going to artificial minds and e on is playing i think a very important role in doing some early pioneering experiments and developments to try to level the playing field of the future not just artificial minds playing on all of this trillions of dollars of infra and data centers but enabling emulations and if you will uploads of human minds non human animal minds at the moment starting with fruit flies and aspirationally working towards mice and humans and getting getting the brains of a variety of human and non human animals to operate in the cloud so walk us through the fruit fly experiment why it was so exciting for you it's really interesting if you think back remember back with the launch of chatgpt that was just a side experiment that wasn't the main focus of openai at the time some have characterized that as an unhobbling if you will gpt three the leadership of openai thought at the time was sort of the main entree if you will similarly within eon i don't to want
Boom. And that's where a lot of our fossil fuels come from. Today, I believe data centers only consume a single digit percent of US electricity. So you're looking at 0.6% of a few percent is like the actual impact. So the short term impact of high oil prices should be very limited. On AI building out new capacity is that, that said, it's worth noting, I think, I think it's something like Qatar produces something like 20% of the world's liquefied natural gas. Right. So we're not for us. Yeah, not, not for us. So we're in a good spot. But this is going to impact the data center build outs or active data centers across the rest of the world if this continues. Yeah. So when you're talking about building new capacity, building new data centers, oil is a little bit more involved. So diesel powers trucks, trains, boats, barges, generators, pumps, compressors, excavators and tons more construction equipment. Petroleum is also broadly used for plastics, polyurethane and solvents that all work their way into the data center supply chain. The biggest problem is delaying already tight schedules because of narrowly available components going out of stock. So the price of oil goes up. There's one marginal factory that can't produce one ingredient that goes into the rack and that slows things down. You have to wait a week while you find another supplier. That stuff can add up to just a little bit of a delay. This happened during COVID and the AI industry was already experiencing something similar with transformers. And so you don't want products getting stuck in transit or going out of stock. But the bigger problem and the one that people should be talking about, and I think you were debating with Dan Primack at Axios about this is macroeconomics. So higher prices, higher oil prices lead to higher inflation. If the Fed has to raise rates to control inflation, capital formation for mega projects gets a lot harder. So JL has this estimate. The next 100 gigawatts of data center capacity could require about $870 billion of new debt financing. And so using this rough number, every extra 50 basis points of borrowing cost on 870 billion is 4.35 billion in annual interest expense. And so that's, I don't know if that's a huge squeeze on cash flow, but it's certainly a squeeze on cash flow. If you're a hyperscaler and you're going gig long AI Capex now and you're like, yeah, we're going to do 200 billion this year, we're going to do 500 billion next year or whatever across a couple companies. You start looking at interest expense bills that are already huge, but then they're going to go up by 5 billion or up by a full extra billion just with a 50 basis point increase, half a percent. So half a percent change in the fed funds rate can echo through. Even a quarter percent could probably, could probably be amplified into a multi billion dollar cost line on your balance sheet or on your income statement as you're servicing that debt. Overall, higher oil prices do make the AI rollout harder, but not really for existing AI capacity. I wouldn't expect token pricing to change based on the price of oil. But there's this bigger question of macroeconomic resilience during a time where our largest tech companies are digging through the couch cushions to find every penny to win the future. Middle Eastern investors pulling back. I agree with you. I do think that this is a significant risk because if they pull back on big, big investments makes these mega rounds more difficult. Retail investors could fly to oil stocks or defense stocks. They think that this is like going to be a thing for months and months and months. It could all flip if there's a quick resolution though, which is what I'm hoping for. But if things continue, the industry walks an even tighter tightrope. So yeah, I mean the big question right now is the hundreds of billions of dollars of of sovereign AI projects in the Middle east, right? I think a lot of those people are going to be like, do we want to send, you know, billions of dollars of GPUs over there? And then also the money coming here is another thing where you might want to spend it elsewhere quickly. Let me tell you about TurboPuffer, serverless vector and full text search built from first principles and object storage. Fast 10x cheaper and extra.
That's very exciting. This is a new Sci Fi corner. I love it. Every Sci Fi trope everywhere, all at once. Time travel. Good framing. Is time travel on the table? We don't know yet. So it's an interesting question. We don't know whether the physics of our universe are compatible with, I think what most folks would construe as time travel. There are versions of temporal nonlocality that are consistent, but with the physics that we have right now, but not in a useful way that would be worth filming a Sci fi movie over. What about faster than light travel? We don't know yet. Same problem in some sense. Exactly the same problem as time travel. We don't know whether the physics of our universe will ultimately allow that, because it seems like if you don't get ftl, it's going to be a boring travel to go to another solar system. I will say this. I think odds are pretty good that AI in the next few years will tell us whether the laws of physics of our universe are compatible with that or not and help us solve this. One of the reasons why I helped found physical superintelligence to discover any of such laws. Very cool. Well, thank you so much for taking the time to come chat with us. Yeah, great to meet you. Congratulations on the progress. Fascinating stuff. Thank you.
Interesting architecturally, what does the neural network look like that you've built for the fly? Is it similar to a large language model like transformer based architecture or are you a beneficiary is all you need. Totally unlike the transformer. Totally unlike the transformer. So transformer architecture, if you subtract off the encoder and decoder layers looks comes in many variants but the most vanilla variant looks like alternating linear and attention layers. So totally unlike the transformer architecture. At least if you look at it it looks like more like a little bit more like a graph neural network. But really that, that's such an AI way of framing it. It looks like a leaky integrate and fire lif model of that that we've had for, for decades from the neuroscience world. So it's, it's just a graph of nodes that have leaky and terrain fire dynamics and they're firing at each other. So what does that tell you about the nature of synthetic human intelligence? Do you think that there is a path to human level AGI that does not involve the transformer and instead falls along your path and your architecture? Oh sure. Well, first of all, I would argue We've had AGI for at least six years now. Five to six years at the very latest. Since summer of 2020 when OpenAI published their language models or FUS shot learners paper. Yeah, GPT3. I would argue that was an AGI. So I, I think there are many ways. My bet is there are many ways to achieve generality of intelligence. I published a paper a number of years ago arguing that not only is intelligent behavior a general process, I, I went further and argued that it is a general physical process that you can even formulate intelligent behavior in pure thermodynamic terms. So my bet is intelligence is this very, very general effect and lots of ways to implement it. Interesting. You've mentioned AGI six years ago, singularity.
Not have been possible a few years ago. So I spend a lot of my time thinking about what the post singularity state of the world looks like and how to smooth that out and bring it here sooner. Yeah. What does the post singularity world look like where you are able to upload yourself? I assume that feels like there's a copy of Alex in the cloud and then you still exist and not in the cloud. How do you think that that interfaces in my mind? If we wind up having the copy question come about, that's almost a failure mode. Speaking for myself, I don't want a copy of myself. I want myself. And Hans Moravec and others have written about this. What I would like to see in the space and what Ian is working toward, at least aspirationally, is this idea of a continuous transfer of consciousness so that it really is you. It's not a copy of you. It's not a low fidelity facsimile of you. It should be a better expanded version of you. That's still you. And whether that looks like moving from what Eon's doing right now with fruit flies to maybe replacing a single cell in a brain at a time, invasively or non invasively, with a substrate independent or substrate migrated implementation, Some variant of that, I suspect is where all of this will go so that you never have to worry, is it really myself or not? It will by construction be yourself. And I'm not sure of the precise timetable, But I think five, 10 years from now, I think the world will see marked progress on the problem of not just whole brain emulation, but also transfer of human intelligence to new substrates. That's very exciting. This is a new Sci Fi corner. I love it. I've read sci fi trope everywhere, all at once. Time travel. Is time travel on the table? We don't know yet. So it's an interesting question. We don't know whether the physics of our universe are compatible with, I think what most folks would construe as time travel. There are versions of temporal nonlocality that are consistent with the physics that we have right now, but not in a useful way that would be worth filming a sci fi movie over. What about faster than light travel? We don't know yet. Same problem in some sense. Exactly the same problem as time travel. We don't know whether the physics of our universe will ultimately allow that because it seems like if you don't get ftl, there's.
And if you live in la, probably still be able to go to the beach. I think so. So that's. But maybe not. Maybe everything's. I did have a funny moment yesterday. I was telling you, but I'll share it again. I was talking with my. My car detailer who's become a dear friend, and he was telling me that the people in my area cancel their appointments at a much higher rate than the people in the rest of LA. And he was talking about running analytics iMessage. And I was like, oh, how are. How do you do. How do you do that? And he's like, he's fully running openclaw. So he's now running, using openclaw to run his car detailing business. He's like, yeah, just always running in the background. To him, he was like, it's some combination of having like an EA or a CEO, like helping me manage the calendar, helping with comms, helping me make smarter decisions on routing and scheduling and all these things. So when you book a detailing, meaning that this gentleman is coming to your house and cleaning your car, you send an imessage to him. Correct. And that is what is so interesting about openclaw is that there were probably a thousand companies that were like, we're vertical SaaS for auto detailing. You got to do everything over email. And they just got eaten alive because people use imessage. And imessage has been this closed ecosystem that private companies could never access before. But it makes so much sense for him to say, hey, run through all my imessages and if I'm talking to my friends about the F1 race, just discard that. But if I'm talking to a client about potentially booking a detailing, put that in a spreadsheet, organize that, and then run some analytics and tell me what the patterns are. And that's just something that should have been SaaS for a decade. But Apple has a walled garden that has now been smashed down by OpenClaw. And I think it's really cool. And I think that that's what we're seeing the first, like use cases for OpenClaw. It's like things that could not happen for business reasons, for logical reasons, sometimes for legal reasons. If you're talking about like downloading illegal films, like the Napster analogy, But a lot of times it's just, it's just Apple didn't want to do a deal to integrate iMessage into Salesforce. The way to break down the walled garden was just to have a digital guy. Basically. Yeah. Who just combs over all this. So is that a new type of software that should have existed that now is unlocked? Is it disruptive? Like, it's not like he. It's not like he stopped using his previous CRM, probably didn't have anything. So now he has something. Yeah, he's pretty tech forward. He's tried every single SaaS tool that there is. But at the end of the day, having a digital guy that has access to all of your tools is just a much better experience. And so I want to know what his bill is. He said he spent a decent amount. I want to know what it is this month and then I want to know what it is in a year. Because sometimes the maintenance of these systems can be really big. And then there's a question about are we going to see per token deflation or are we going to see prices? Are prices being artificially held low because of competitive dynamics? So in a year, will he be spending more or less? Will he still be getting the value? These are all interesting questions, but I'm just excited that a digital product now exists that could not or didn't exist before. And that feels incredibly positive and cool. So very.
And podcast. Have you been noticing that it's been hotter in Los Angeles? I have. Downtown Los Angeles is forecasted to approach a hundred degrees Fahrenheit on Thursday and Friday in March, which is why not good. We're gonna do the weather. New segment for you today. We're doing the weather on tvpn. We have our very own Ben. We have Ben. Hello, guys, how are you? We're doing great. Tell us about the weather. What's happening? Well, I want to start off by saying, as you can see, the weather today for the low today is going to be 75 degrees Fahrenheit up there, high of 100 degrees Fahrenheit down there. But there's something I actually wanted to point out that I saw and I thought that was quite interesting. As you can see up here, there's a localized low pressure area up there and a localized high pressure area right down there. If you can see that right there, does that mean rain? The issue that's not normally an issue, not a cause for concern, and it's not very common for this time of the year. However, today, later in the afternoon, these two areas are going to collide. They're going to hit each other. Really? And what that's going to cause is a barometric pressure inversion. Okay, it might sound a little bit scary, but I guarantee there's no cause for concern. All that means is that hot air rushing in from the west is going to collide with that cold air rushing in from the east and it's going to cause a bunch of turbulence in the sky, moving all the airwaves around and oscillations in the sky. However, I want to add one more point. A byproduct of this effect is that all that humidity that dropped after that hot air moved to the bottom is going to raise up because the water cycle, you know, evaporation and stuff, it's going to raise up into the sky, into those clouds, going to cause big clouds in the sky and eventually all that water is going to fall down onto the ground. We're going to big rain later in the afternoon. It's actually going to rain in L. A? Yeah. It might sound crazy, but I just want it for all you guys at home. I definitely try to step outside with a jacket today, maybe a hoodie just in case the rain comes. Don't try to be a big shot. I'm fact checking you right now. And the weather app says it's going to be sunny all week. Is this just complete fake news? My team? Those are the numbers my team gave me. This is complete fake news. No, no, no. This is the fakest news I've ever heard. All that literally. The transatlantic current. John. John, I'm. I don't trust your app. I trust the weather. Apple says it's not gonna rain the entire month. There's zero chance of rain. Did you look at the transatlantic current? No. The transatlantic current. Precipitation. Zero inches today. That's putting you in the streets tomorrow. Zero inches on Wednesday. John, you're really gonna. You're really gonna trust. Really going to trust an application that was probably vibe coded yesterday? Over. Ben, who's doing the weather. How did he get here? What happened here? Yeah, no, if you look right there, you can see a localized high pressure area. That's a localized. I think this is an overeager weatherman who's just looking for drama in the most boring weather market in America, which is Los Angeles. I think. I think this is the most important story. Look at the world. This is ridiculous. You guys can look at the jet streams. They're coming in from the west. I don't want to hear any more mumbo jumbo about jet streams. Get out of here. You're done. Great work, Ben. Thank you, Ben. The fake news weather will return. Maybe never. Who knows? Let me tell you about Restream. One live stream.
Accounts to move money, fight fraud, and improve lending now with AI. And let me also tell you about Cognition. They're the makers of Devon, the AI software engineer. Crush your backlog with your personal AI engineering team. So the parent company of Four Loko, the canned alcoholic beverage that became a college campus sensation in the late 2000 before being reformulated under regulatory pressure, is exploring. I cannot believe they nerfed 4 loca. It was too powerful. It was too powerful. Like, you can just imagine the trajectory of the United States if it hadn't been nerfed and then straight downward. Really? Yes. It was so bad. It was so. I thought people were having, like, heart attacks. Yes, it was terrible. So 4 loko was a 24 ounce can. So twice the size of a normal alcoholic beverage. Trey, the birthday boy, says it was amazing. It was amazing. Okay, now we have buzz balls. That's true. So Four Loko, the original formulation, was twice as big as a normal can of Bud Light or Miller Lite or Coors Light, your normal beer, something that you would grab. So it was two drinks that way. And then instead of it being 3.2% alcohol or 4.5% alcohol, it was like 10% alcohol. It was the strength of wine almost. And so a single Four Loko was like four or five beers in one can. And then they also added like 200 milligrams of caffeine. So you would become incredibly intoxicated and inebriated, but then also incredibly stimulated from all the caffeine. And that spelled doom for many people. People would be very high functioning, but completely inebriated and discombobulated. And so they would get into all sorts of trouble, whether or not it was behind a motor vehicle or the discombobulator. It was the discombobulator. It was also wildly illegal from an FD perspective. You can't mix alcohol and caffeine in a single product. That's just a rule that they chose not to follow. And they figured out how to make it. And then the FDA eventually gave them a warning letter and they pulled the caffeine out. And that, of course, killed the viral sensation that was for Loko. But apparently the company continued to sell products that were compliant with the current FDA regulation. Nerf. That's nerfed. The Nerf for loco continued to sell. Somehow they're looking to and potentially will capture somewhere around 400 million on the sale. The brand value. It's hilarious that they're working with J.P. morgan on this. That's wild. So Fusion Products is the owner of the brand. They're working with JP Morgan on the sale process. The potential sale underscores how ready to drink beverages have emerged as a growth category in an otherwise sluggish alcohol market. U.S. beer, wine and spirit sales decline declined in 2025. We've talked to a number of guests about the declining alcohol sales as folks move over to more functional beverages. Here's what the CMO had to say. For over a decade, our sales have been the leading flavored malted beverage. Have have been leading the flavored malted beverage market by embracing bold innovation, unconventional marketing and a risk taking attitude that delivers results year after year. I guess they kept building. They also own the parent company also owns Pirate Water, Pirate Water, Basico Tequila and just Earthquake. Earthquake. Earthquake seems like something you definitely want from the makers of Four Loko. Earthquake Pirate Water is a wild name for a brand too. I mean it's a distinctive can. It looks remarkable. It jumps off the shelves. When the Four Loko Nerf went into effect and they pulled them from the shelves because before they reformulated without the caffeine, they had to pull them from the shelves entirely. They were a hot commodity and folks would go around stockpiling them before they were banned. And they were one of the most guarded resources in all of college campuses. Wow. Earthquake is just a can that's just 10% alcohol. Wow. Do not recommend. Very risky. Never quake. Anyway, in 2010, the FDA sent warning letters to caffeinated alcohol beverage makers, including Four Loko, saying that the caffeine in the their products constitute an unsafe food additive. So the FDA has specific categories for regulation around what is an alcoholic beverage, what is a caffeinated beverage. And they never the two shall mix. So they don't let you put these two things together. Of course, vodka Red Bull was the precursor to this where folks would order a Red Bull and then they would put vodka in it, I think, but at a bar which is made at the time of sale. And so that was legal, I believe. And this might be fake news, but I remember hearing a story that at one point Red Bull.
Uav online. Glaze. Double glaze. Triple glaze. Double kill. Five. Kill. Roar. Wins. Team deathmatch. We are experts. Triple blades. Let's just roll, right? Market clearing order inbound. You're surrounded by journalists all your. Strike1, Strike 2, Activate. Go golden retriever mode. Market clearing order inbound. 5. I see multiple journalists on the horizon. Founder, You're watching TVPN. Today is Monday, March 9th. We are live from the TVPN Ultra Dome Temple of Technology, Fortress of Finance, ramp.com Time is money, save cards, bill pay, accounting and a whole lot more all in one place. We have a fantastic show. We were able to get Alex Epstein to come down to the Ultradome in person on short notice to get us up to speed on what's happening in the oil markets. Of course, oil is all. It's gushing all over the timeline, all over the Financial Times, all over the Wall Street Journal as part of the. And Alex had a feeling that oil would be important in the future when he wrote the book Fossil Future. Yes. Which we have right here. Which if you haven't given it a read, go pick a copy up and stay tuned for him joining us at 1:10 today. Let's pull up the linear lineup and tell you who we have coming on the show today. Linear, of course, is the system for modern software development. 70% of enterprise workspaces on LINEAR are using agents. Now we have Dr. Alex Wisner Gross over on eonsystems. He went very viral over the weekend for connecting a biologically derived fruit fly connectome to a Mujukou physics simulated body. And we're going to be talking all about simulating fruit flies and what will come after that. And we're also going over Microsoft copilot cowork Julian, which we're going to call Coco Coco. Okay, Disney might like a word. They have IP in that space in particular, but we'll see. Let's see. They have it in the enterprise. I don't think so. Julian's joining from Sequoia Capital. Owen McCabe from Seattle from Intercom. Our partner is joining because they just raised a massive clean 200. And Michelle Voles is launching Pax VC. She left Andreessen Horowitz I think about a year ago and has been is now launching a new early stage fund. Anyway, let's go through the timeline and see how people are processing the news of the Strait of Hormuz and the oil price spikes. Fox News this morning they were saying just basically their advice to the captains the man up and just send it yeah, send it. That seems crazy. Thereby, please stay safe. I'll read the actual quote. A guy, Brian said, if you want to diminish the Iranian threat, if you want to make sure this ends with complete capitulation, show some guts and go through that straight. That seems very, very risky right now. Stay safe out there. If you are on a shipping vessel, do whatever is responsible. There are some crazy twists happening. So apparently ships in the Gulf are declaring themselves as Chinese vessels to dodge attack. This is from the Financial times. At least 10 vessels have changed transponder messages in an apparent attempt to avoid becoming targets. A clutch of vessels trapped in the Gulf under enemy fire are adopting a tried and tested ruse to avoid attacks. They're changing flags, using transponders to declare themselves to be Chinese. There's always been a very odd like, just like tug of war between how ships identify themselves because often for tax reasons, they're bought in one country, operated by individuals from another company, but they fly a different flag to be able to go from one place to another. And it's all based on like the port systems. I don't fully understand it, but it is very interesting. At least 10 ships over the past week have altered their dest signal to read Chinese owner. All Chinese crew or Chinese crew on board. About 1,000 ships are currently shut inside the Gulf and its immediate surroundings with a cumulative value of $25 billion. And I don't know if you've seen some of the maps that show the Strait where it looks like nothing is actually moving through. Yeah, I think in actuality a number of the ships are actually turning off their transponders, so you can't see the movement because they're basically moving a little bit going through the Strait and then turning back on again. Well, there are plenty of outlets that are covering the Iran conflict in detail. We of course will be focused on the business and technology impacts. I wrote about what oil prices mean for the AI buildout in data centers. It's just sort of interesting to dig into the deeper supply chain of artificial intelligence. But there are some posts that we should go through around the oil story. So Crude Oil is five standard deviations above its 50 day moving average. Statistically speaking, this occurs every 9,500 years. So the last time would have been about 6,000 years before Moses parted the Red Sea. Imagine what that did to shipping in the area. Fanciful. What else is going on? The Kobsi letter has been posting a lot about oil prices. US oil prices are now expected to rise above $100 a barrel this month that already happened. This was posted on March 6 with markets pricing a 66% chance that it happens. The last time oil prices were above 100 was July of 2022. Let's pull up this clip from Landman. I haven't seen Landman. Have you watched it? Is it good? Let's. You want oil to live above 60, but below 90. And don't get me wrong, we're still printing money at 90, but gas gets up a bit. Little 350 a gallon. It starts to pinch. It hits 100. Every product in America has to readjust its price. $78 a barrel. That's about perfect. You know, it brings enough profit to keep exploring but don't sting as much at the pump. Let's course here in California, I mean, they tax the out of it out there. It could be $45 a barrel and it's still $4 a pump. I don't know how those slimes do it out there. It's movie day 2020. A barrel of oil was worthless. This place became a ghost town and nobody's immune. Kids have to quit college. Trucks get sold or repoed. We can pause it now. We got the gist of it. Yeah. Well, this is the largest. How old were you when you discovered that gas is really expensive in California? Not just because there's a lot of demand for it. I mean, I discovered it when I was filling up in Montana and it was. Was like $2 a gallon and it's like $5 a gallon here. That was pretty wild. The first major gas price shock that I noticed was Hurricane Katrina. In my life, I think it was in high school because I was too young. I was talking about just the fact realizing that it's so expensive because of the taxes that California puts on it. Yeah, because it's nice to drive around here, put the top down, the weather's good. So they're like, yeah, you're gonna pay. You're gonna pay more for the joys of driving an internal combustion engine car. Now, the price in California is aggressive. I don't know the structure of the prices though. Is it percentage based or fixed? Because that affects how much the price will move based on oil price shocks. Because if the price per barrel doubles but the tax is flat, you don't feel it as much here as you do other places. I don't know. Anyway, I'm pulling it up. But you can run through this. The largest supply shock by a factor of four. So the Hormuz blockade, which is current 20 million barrels were lost in supply. The Iranian Revolution in 1978 was 5.6 million. The Yom Kippur War embargoes in 73 was 4.4. The Iraq Kuwait War was 4.3 in 1990. The Iran Iraq War in 1980 was 4.0 and the Ukraine Russia invasion in 2022, which is the last time that oil spiked over $100 a barrel was 1 to 3. So an absolutely huge supply shock and I'm sure it will have a lot of implications all over the economy with triple digit prices. Here's what's going to happen now, says policy Tensor markets will scream when they open tomorrow. VIX will surge to levels beyond what we saw in April. The sell off will continue for some time as intermediate as intermed shed risk. And the markets are red. They have been screaming today. The VIX futures curve already has inverted bid up by dealers looking for insurance. This predicts a massive sell off. The pressure on this captured White House now, the pressure on this captured White House now beings in earnest. That's sort of oddly. Anyone can tell them that if this persists for very long it will destroy the Trump presidency and gut the GOP for a generation. The controlling factor here, as I have told you over and over again, is that the United States does not have the military means to reopen Hormuz. There's no military solution in sight. This means that not only does Iran have the strategic upper hand now, it means that Iran enjoys the unambiguous strategic advantage. All they need to do is keep the thing closed until he capitulates. I put my neck out bar to call this in advance and someone told him yesterday I was in the minority, perhaps even a minority of one. No longer. I was correct just calling a shot. The blob heads and scribes were incorrect in their assessment of the strategic situation and now markets will price to reflect reality. And we briefly touched on Scott Sumner's blog post on Substack at the end of Thursday's show, maybe Friday's show. We didn't get a chance to read it. I actually read through it earlier today and it's pretty interesting. It basically makes the case. It has a very controversial title Freak out. And the thought is that he's doom posting. Maybe it's about AI, maybe it's about this particular conflict. He is more just reflecting on this dynamic between when the market freaks out it acts as a moderating effect to policy. And so he gives a number of examples around like tariffs or more specifically the admin. Yeah, yeah, like the tariffs caused this massive circuit breaker, 5% sell off in the market. And then that was internalized and very quickly adjusted. And there were a whole bunch of different carve outs to sort of like create soft landings. And so he's actually. And what he's getting at is that after the fact, a lot of people say, look, you didn't need to freak out because the taco happened. Trump always chickens out. The actual proposed policy effect didn't go into effect. And his point is that, well, it's precisely because people freaked out that it didn't go into effect. So freaking out is good in Scott Sumner's mind at least. Anyway, I love this. Yeah, we have a solution. We have a solution. If you're feeling the pain at the pump pivot and get a horse at oil at 110, the urban horse is the only option. Pulling up to the gas station on a horse is truly elite. I do want to know what's the TCO on a horse with the food and the stables versus, you know, just keeping a. What is that, a Ford Taurus in your. I don't know, that's something else in your garage. The horse really mogs at the gas station in particular. Right. Because it's just making everyone feel so stupid. Yeah, we got to go back. One horsepower is all you need. Nick Carter says seeing a lot of non process trusting Pannekins on the tl and then he followed up by saying, that's what I thought because the oil prices spiked up and then they fell down. And we did not get $120 a barrel oil. We got exactly 100, which was still not like it's way, way, way too early to celebrate or anything. Yeah, exactly. There's an interesting chart here, a really shocking chart from Oliver Grob. Unreal numbers about Germany's nuclear power generation. So if you scroll down, look at this curve. This is truly the Bell curve meme or something like that. JP Morgan estimates that had Germany not phased out nuclear power, the country would have generated 50% less electricity from fossil fuels and 84% less electricity from natural gas. In 2024, electricity prices in Germany would have been around 25% lower, and the country would have imported half as much electricity and just complete rise and fall of nuclear power generation in Germany. One of the craziest graphs. You don't see graphs like that very often in new technologies. Typically you see S curves or you see exponentials. No one considers. The models get better and then they get dumber. That's certainly the funniest Outcome? Yeah, very, very rough. Goldman Sachs sent a note to investors saying if oil prices increase by $10 and remain elevated from for three months, US year over year, headline CPI inflation would likely rise from 2.4% in January to 3% in May. Those are small numbers, but we're looking at an oil price increase of maybe $30, $40 over the baseline, potentially higher. We don't know where oil's going to land. And so you have this weird tug of war right now with the Fed where if, if oil prices go up, inflation goes up. The Fed has a mandate to curb inflation. That means higher interest rates. At a time when the labor market is shedding jobs, you would expect a Fed rate cut or a lot of people are optimistic about a rate cut. Trump certainly wants rate cuts. But if inflation is climbing, there's really no solution other than keeping rates higher or even raising them further. So a real jam in terms of federal monetary policy, Fed monetary policy. Here's Art Cashin at the opening bell. This is a historical video. When was this? This was a long time ago. Let's play this clip. This may be it. First let me start out Muratori te salutamus essay. And you know that's the gladiator salute. We who are about to die salute you. So it's gonna be a tough morning. This may be it. You are about to die salute you. Insane, insane aura for. You know, CNBC really doesn't get enough credit for being so innovative in terms of broadcasting and entertainment. Really some of the greatest clips. I love one of the top comments. This is an old clip. Yeah, obviously Brother 4 ADP. It's probably from the 90s. So Zero Hedge says, you know, with oil at 111, total panic and memetic Sisyphus shares. A clip says half of this site for the last week has sounded like this. Let's producers are telling me there is breaking news. The Asian financial markets have just opened to a huge sell off and we're gonna switch to that story right now. Good, I'm glad I'm here. Your thoughts, Tracy Jordan on how this is gonna impact Wall Street. Larry, I'm not an expert but I do have a strong opinion. New York as we know it will no longer exist tomorrow. Producers are telling me there is more. Is that from 30 Rock? That's so good. That's so funny. Oil came way in from its overnight highs, says Joe Weisenthal. The quote post is Chris Paul hits a huge three to cut down the lead to 42. Absolutely, absolutely brutal. So let me Tell you about Gusto, the unified platform for payroll benefits and hr built to evolve with modern small and medium sized businesses. Stop making fun of us. We're somewhat of a modern business. Small and modern. Small and modern. Smodern. Let me also tell you about CrowdStrike. Your business is AI. Their business is securing it. CrowdStrike secures AI and stops breaches. So we got to give you a shout out to George Kurtz for the Mercedes performance yesterday. Oh yeah. 1, 2, 1, 2. Really good. George Kurtz. Would you look at that. Take the stake in the team, suddenly they're at the top. Yeah, he's on to something. Well, my essay this morning was titled why is no one Talking about Oil? Of course everyone is talking about oil. Oil discourse gushed onto the timeline this weekend. Crude prices spiked to nearly $120 a barrel as a broadening war in Iran threatens both transportation routes and production. The geopolitical and economic analyses are flowing, but what does this mean for AGI timelines? And a lot of people in the AI world are sort of tuning all of this out because they see recursive self improvement, AGI, asi, the build out as more important. And I just wanted to sort of reality check the AI supply chain to understand how does oil actually affect data center construction, AI production, token pricing? Like is there any effect? My conclusion was that it's very moderate, but there are some interesting effects in the financial markets that are probably the bigger takeaway. But it's still interesting to hear about. Like yes, oil actually does is used in the production of AI at least a little bit. So power has been at the forefront of the AI pushback. Like everyone's worried about these local energy prices, these electricity prices increasing near the data centers. But, and it's been, it's become like a political issue. But pain at the pump might become a bigger story as gasoline prices spike. And, and that has been pain at the pump. Oh, it's been the most tangible sign of inflation. It moves so quickly, one jitter in the economy and it's a huge component. People on the coast, people in tech don't have a good sense for this. Right. If gas, gas. For a lot of people, gas could quadruple and they wouldn't really notice it. No, no. But if you actually look into the average American, how gasoline fits into their budget, it's a meaningful component of their monthly budget. So they feel it super intensely. It's variable cost. There are so many different ways where a lot of Americans go on driving vacations. That obviously is Directly impacted by gas prices. And then also just psychologically, there's something about filling up at the gas tank where you see the number ticking up and you're doing that on a every week basis or so that it's just so visceral. It's this thing that you have to stop and then go experience and watch the money flow out of your account, like in real time. It's very. It's very visual, it's very interactive. Yeah, I remember, I remember I must have been probably 18 at this point where I would just go. And I would like for a long time. I just put. I'd like, you know, prepay for a certain amount. Yeah, I got 20 bucks. Let's see how much I got. And I felt like, really like, I felt like the king of the castle. I just put my card down and let it run up. Yeah, yeah, yeah, yeah. No, no, totally. So pain of the pump has been like a big, broad economic issue. It's so visceral. Like, the actual phrase has a lot of power in politics and in like attack ads, like your gas prices are going up under this person. That's why you got to vote for me. This is. This is common. But in AI circles, the discussion's been much more focused on RSI now is the new acronym that everyone's focusing on, not AGI. AGI is here. We know artificial general intelligence. We pass the Turing test. The models can do things generally intelligently, but can they recursively self improve? Are they rsi? Are we in RSI now? Is it coming? Is this going to be a fast takeoff? Is this going to be a slow takeoff? Well, something's taken off. Dylan Patel said being in SF is like being in Wuhan right before the pandemic. Something is happening. It's going to hit everywhere, but so few people know it. So he's sort of echoing that something big is happening mentality. This idea of like recursive self improvement being AGI pilled. A lot of smart people agree there's definitely something happening. There's still this question of like, what is sticky? What? Where will diffusion take longer? What will be sort of AGI resistant or ASI resistant for a variety of reasons, even if they're completely irrational and you could get a better thing from a computer. But people like the human version or whatever. The irony is that George Hot's hitting the timeline to raise money makes me more bullish on acceleration because if not, he's obviously not historically been a huge fan of venture capital. Well, he's not raising from traditional VCs. He said he specifically said he wants to raise from like basically like friends and family type of round. He does want 20 million. We have this in the timeline somewhere. Where is this post from George? I'll pull it up. If you want to keep Tiny Corp. He said that he's raising $20 million to buy a building. He's going to buy a building. Here it is. So he said this is from Tiny Corp. The makers of Tiny Grad and the Tiny box. If Tiny Corp. Was raising $20 million at $200 million valuation, who'd be interested? Business model is basically this. Buy an $11.5 million building with 5 megawatts of power. Link in our Discord, wait for AMD to launch the rDNA 596 gigabyte cards mid-2027 pre order 3000 cards. Hopefully we can negotiate for $2500 each. Build $520,000 tiny boxes with 6 of the cards in each box run all the Chinese LLMs make $600,000 per month revenue selling tokens on open router. Market depth is there. This is 1% of open router improvements to Tiny Grad yield revenue improvements Due to how power is priced in Oregon. It's only like $50,000 for the electric bill before the 4 megawatts before they price for peak not usage. We get like 3c kilowatt hour power.3 cents per kilowatt hour. We can also make $100,000 per month leasing colocation space to comma building and cards paid off in three years. Max investment made back low risk of being undercut since we're using consumer GPUs and running the cheapest colo you can believe. If someone chill wants in, I do it. I'm not gonna fake hype. I'm not gonna hype fake tech. But demand for tokens is gonna skyrocket. Skyrocket. Look at the open claw install numbers. With crazy good optimizations, we could potentially get 3x more from the machines. And we have electricity 3 for 3x more machines, 5.4 revenue per month. Then continue to scale from there. Custom chips, et cetera. He's starting a Neo cloud. Are you starting? Yeah. He's going to be serving tokens. So that was incredibly bullish. Demand is definitely there. At the same time, he sort of took the other side of like this is different. This is crazy. He says there's no AGI. There's no magic threshold. You guys see auto research? Change the random seed from 42 to 137 and OMG it's AGI, it's over. Yet you critique the junior engineer for the same stuff. The cost of development is falling. Overpaid engineering struggles to compete. That's the story. And he says humanity has been recursively self improving for centuries, so everything new is old. I don't know. Obviously have a huge amount of respect for George, but at the same time saying there is no artificial general intelligence when you're also arguing that engineers are struggling to compete with new technology that is replacing them. This. How is this not artificial intelligence competing with. Yeah, so. So I think he is. He is agreeing with that, but it's that it's that there is an immense amount of. Of desire for this binary moment. This is the singular. This is AGI, this is asi. The RSI is here. This thing is happening right now. And there's before and after, and everything has changed. And he just doesn't see it that way. I think. I think he sees it much more like the Internet, the mobile phone, like other technologies that have been rolled out. Electricity. Yes, there is like a before and after, but you can only really define the period by maybe a decade, and you need a few decades to understand that moment. It's very hard to go back. And, you know, there is like the iPhone moment and there is like, you know, the first launch of, I don't know, aol. I don't even know what the iPhone moment of the Internet was, just because it was sort of a slow rollout anyway. Live GPU clusters. Where's Jordi going? Oh, you're moving the goalposts. Okay, Jordy's moving the goalposts. Where are you moving the goalposts to? Just over here. Okay, but why? Is there a reason? I mean, they have to be moved. George is moving the goalposts. Oh, okay, okay, okay. He's moving the goalposts. So I said, you know, it's clear that the AI industry continues to grow and continues to need more and more power and compute, as we've seen from George Hotz's new project. That means large data center campuses. But if they're not in random office buildings that George is picking up on the cheap, they're probably going to be built with construction equipment. So what does this mean? They don't just drop from the heavens. They require building, which requires oil. But how much oil? And is oil a serious. Is a serious oil shock enough to impact the AI buildout in a meaningful way? Spoiler alert. Basically, no. But live GPU clusters in the United States do not use Much oil directly. Only 0.6% of U.S. electricity in 2024 came from petroleum. We're much, much more dependent on natural gas. Something like 42% of US electricity is natural gas. And so America ramped up natural gas production significantly over the past two decades. A lot of that was in reaction to the wars in the Middle East. Hey, we need to be less dependent on foreign oil. We need to be energy independent. And so you have the fracking boom, the natural gas boom, and that's where a lot of our fossil fuels come from. Today, I believe data centers only consume a single digit percent of US electricity. So you're looking at 0.6% of a few percent is like the actual impact. So the short term impact of high oil prices should be very limited on AI building out new capacity is. Yeah, that said, it's worth noting, I think, I think it's something like Qatar produces something like 20% of the world's liquefied natural gas. Right. So we're not for us. Yeah, not, not for us. So we're in a good spot. But this is going to impact the data center build outs or active data centers across the rest of the world if this continues. Yeah. So when you're talking about building new capacity, building new data centers, oil is a little bit more involved. So diesel powers trucks, trains, boats, barges, generators, pumps, compressors, excavators and tons more construction equipment. Petroleum is also broadly used for plastics, polyurethane and solvents that all work their way into the data center supply chain. The biggest problem is delaying already tight schedules because of narrowly available components going out of stock. So the price of oil goes up. There's one marginal factory that can't produce one ingredient that goes into the rack and that slows things down. You have to wait a week while you find another supplier. That stuff can add up to just a little bit of a delay. This happened during COVID and the AI industry was already experiencing something similar with transformers. And so you don't want products getting stuck in transit or going out of stock. But the bigger problem and the one that people should be talking about, and I think you were debating with Dan Primack at Axios about this, is macroeconomics. So higher prices, higher oil prices lead to higher inflation. If the Fed has to raise rates to control inflation, capital formation for megaprojects gets a lot harder. So JL has this estimate. The next 100 gigawatts of data center capacity could require about $870 billion of new debt financing. And so using this rough number every extra 50 basis points of borrowing cost on 870 billion is 4.35 billion in annual interest expense expense. And so that's, I don't know if that's a huge squeeze on cash flow, but it's certainly a squeeze on cash flow. If you're a hyperscaler and you're going gig a long AI capex now and you're like, yeah, we're going to do 200 billion this year, we're going to do 500 billion next year or whatever across a couple companies, you start looking at interest expense bills that are already huge, but then they're going to go up by 5 billion or up by a full more, a full extra billion just with a 50% basis, 50 basis point increase, half a percent. So half a percent change in the fed funds rate can echo through. Even a quarter percent could probably, could probably be amplified into you know, a multibillion dollar like cost line on your, on your balance sheet or on your income statement as you're servicing that debt. Overall, higher oil prices do make the AI rollout harder, but not really for existing AI capacity. I wouldn't expect token pricing to change based on the price of oil. But there is this more, there's this bigger question of macroeconomic resilience during a time where our largest tech companies are digging through the couch cushions to find every penny to win the future. Middle Eastern investors pulling back. I agree with you. I do think that this is a significant risk because if they pull back on big, big investments makes these mega rounds more difficult. Retail investors could fly to oil stocks or defense stocks. They think that this is like going to be a thing for months and months and months. It could all flip if there's a quick resolution though, which is what I'm hoping for. But if things continue, the AI industry walks an even tighter tightrope. Yeah, I mean the big question right now is the hundreds of billions of dollars of sovereign AI projects in the Middle East. Right. I think a lot of those people are going to be like, do we want to send billions of dollars of GPUs over there? And then also the money coming here is another thing where you might want to spend elsewhere quickly. Let me tell you about TurboPuffer, serverless vector and full text search built from first principles and Object storage. Fast 10x cheaper and extremely scalable. And let me also tell you about public investing for those who take it seriously. Stocks, options, bonds, crypto treasuries and more with great customer service. Whatever investing strategy is, we have to say Happy to birthday. Happy birthday to Trey. Happy birthday, Trey. We love you. Excited to have you. Thank you. Of our strongest soldiers. We appreciate you. It's been an honor. Have you been noticing that it's been hotter in Los Angeles? I have. Downtown Los Angeles is forecasted to approach 100 degrees Fahrenheit on Thursday and Friday in March, which is why not good. We're gonna do the weather segment for you today. We're doing the weather on tvpn. We have our very own Ben. We have Ben. Hello, guys, how are you? We're doing great. Tell us about the weather. What's happening? Well, I want to start off by saying, as you can see, the weather today for the low. Low today is going to be 75 degrees Fahrenheit up there, high of 100 degrees Fahrenheit down there. But there's something I actually wanted to point out that I saw and I thought that was quite interesting. As you can see up here, there's a localized low pressure area up there and a localized high pressure area right down there. If you can see that right there, does that mean rain? The issue with that's not normally an issue and not a cause for concern. And it's not very common for this time of the year. However, today, later in the afternoon, these two areas are going to get collide and they're going to hit each other. Really? And what that's going to cause is a barometric pressure inversion. Okay. It might sound a little bit scary, but I guarantee there's no cause for concern. All that means is that hot air rushing in from the west is going to collide with that cold air rushing in from the east and it's going to cause a bunch of turbulence in the sky, moving all the airwaves around and oscillations in the sky. However, I want to add one more point. A byproduct of this effect is that all that humidity that dropped after that hot air moved to the bottom is going to raise up because the water cycle, you know, evaporation and stuff, it's gonna raise up into the sky, into those clouds, gonna cause big clouds in the sky and eventually all that water is gonna fall down onto the ground. We're gonna have big rain later in the afternoon. Wait, it's actually gonna rain in la? Yeah. It might sound crazy, but I just want it for all you guys at home. I'd definitely try to step outside with a jacket today. Maybe a hoodie just in case the rain comes. Don't try to be a big shot. I'M fact checking you right now. And the weather app says it's gonna be sunny all week. Is this just complete fake news? My team. Those are the numbers my team gave me. This is complete fake news. No, no, no. This is the fakest news I've ever heard. All that literally. The transatlantic current. John, John, I'm. I don't trust your app. I trust the weather. Apple says it's not gonna rain the entire month. There's zero chance of rain. Did you look at the transatlantic current? No. The transatlantic current. Precipitation. Zero inches today. That's putting you in the tomorrow. Zero inches on Wednesday. John, you're really gonna. You're really gonna trust. You're really going to trust an application that was probably vibe coded yesterday over Ben, who's doing the weather. How did he get here? What happened here? No, if you look right there, you can see a localized high pressure area. That's a local. I think this is an over eager weatherman who's just looking for drama in the most boring weather market in America, which is Los Angeles. I think. I think this is the most important story. Look at the world. This is ridiculous. You guys can look at the jet streams. They're coming in from the west. I don't want to hear any more mumbo jumbo about jet streams. Get out of here. You're done. Great work, Ben. Thank you, Ben. The fake news weather will return. Maybe never. Who knows? Let me tell you about Restream 1 livestream. 30 plus destinations. If you want to multi stream, go to restream.com and let me also tell you about Gemini 3.1 Pro. With a more capable baseline, it's great for super complex tasks like visualizing difficult concepts, synthesizing data into a single future. We're bringing creative projects. This is the future of the weather. We have a weatherman who gets into a live altercation with one of the other hosts who's just constantly fact checking. It's on site. Next time, Ben, it's on site. If you pull up with some fake news on TVPN's weather report, it's on site. Ben, you crushed it. Good job. So it is. It's not Adam Smith's birthday. It's the 25th anniversary of the publication. 250th. Sorry, 250th anniversary of the publication of the wealth of Nations. Tyler, you have a copy? I have two of our copies here. I don't bring. So. Yeah, usually I have one with me, but I didn't bring mine to work because we have a We have a few here. Yeah, yeah, yeah. Obviously, for those who don't know, give us the pitch for wealth of Nations. Why should people read wealth of nations in 2026? Yeah, I mean, this is kind of the OG, you know, economics book, right. Was. It was like put it into like AI terms. Is this like situational awareness type of. I mean this is like attention is all you need. It was like the chatgpt moment for capitalism. Yes, yes. But I would like to, you know, I have a few quotes I would like to read. Okay. All right, let me open it. Let's see. It's a big book. It's like 800 pages, it's very long and it's really, really dry. I mean, it reads like it was written 250 years ago. Okay, so this is, I think this is kind of one of my favorite quotes. Yeah. Kind of the, you know, seminal. Every individual necessarily labors to render the annual revenue of the society as great as he can. He generally, indeed neither intends to promote public interest nor knows how much he is promoting it. Yes. He intends only his own gain. Yes. And he is in this, as in many other cases, led by an invisible hand, promote and end, which was no part of his intention. Yeah. Rational economic actors, homo economicus, thinking of the value for societies created by rational self interest. Sort of the spin on like greed is good was like, he was like an OG of that thesis. Greed maxer. Yeah. Productive labor, efficient markets, Free markets don't hold gold don't control trade. An important read in 2026. Somewhat heretical. There are many people that disregard that book. Maybe they should not. Maybe they should give it a crack. Green. I think it's evergreen. I think it's evergreen. I mean there was that debate over like Tyler Cowen over the weekend and somebody was taking shots at stubborn attachments for being like, not like, not advancing the discipline of economics. And my pushback to that was that. But I actually think it is incredibly, it's just a fun read. It's like really well written and it's very entertaining and it's very thoughtful. But it's particularly interesting to hear someone sort of just rehash the tenets of liberal democratic capitalism in 2020. I think he read it in 2023, 2022, something like that. But like there are a lot of voices out there who are anti capitalist, anti democratic, anti liberal. And Tyler Cowan is just retreating to like a time tested, time honored economic philosophy that is deliberately not revolutionary. And I thought that it was a great book for that reason. And that's why I recommend stubborn attachments to so many people and so widely. But if you want to crack open the real hard stuff, pick up a copy of the wealth of Nations. Why is one of those copies so much smaller? Is the font size just smaller? Yeah, it's super small. It's also two columns. Oh, two columns. Wow. Yeah. If you can get through that, you're truly elite in the global economy. What does the timeline have to say about the masterwork that is turning 250? It's now needed more than ever, says the Washington Post. On March 9, 1776, four months before the American colonies broke with Britain over the issue of taxation, a little known Scottish thinker published a long, dense book with an unpromising title, Shots Fired An Inquiry into the Nature and Causes of the wealth of nations. 250 years later, Adam Smith is by any objective measure easily the most widely cited and widely quoted economist who ever lived. Astonishingly, his work still frames the central questions we face. Not just about free markets, trade and capitalism, but about the nature of human society itself, and even what it is to be human at all. Smith was born in 1723 in Kirkcaldy, Scotland. Educated at the Universities of Glasgow in Oxford, he initially made his name not as an economist, but as a moral philosopher. His first published book, the Theory of Moral Sentiments, offered a radical theory of how we form moral judgments, radical because it derived from the creation of moral values, not from scripture or divine grace, but from human sympathy and mutual regard. The wealth of nations, as his second work, second major work came to be known, was an extension of that product. The book is not, as sometimes believed, a hymn to greed, a pain to market fundamentalism, or red in the tooth claw capitalism. It was an attempt to understand how a commercial society society could generate prosperity without collapsing into corruption. This 250th anniversary is not a moment for hagiography. It is an opportunity to recover a way of thinking.
So good. That's so funny. Oil came way in from its overnight highs, says Joe Weisenthal. The quote post is Chris Paul hits a huge three to cut down the lead to 42. Absolutely. Absolutely brutal. So let me tell you about Gusto, the unified platform for payroll, benefits and hr. Built to evolve with modern, with modern small and medium sized businesses. Stop making fun of me. We're somewhat of a modern business. Small and modern. Small and modern. Modern. Let me also tell you about CrowdStrike. Your business is AI. Their business is securing it. CrowdStrike secures AI and stops breaches. So we got to give you a shout out to George Kurtz for the Mercedes performance yesterday. Oh yeah. 1, 2, 1, 2. Really good. George Kurtz. Would you look at that? Take the stake in the team. Suddenly they're at the top. Yeah, he's on to something. Well, the, my essay this morning was titled why is no one Talking about Oil? Of course everyone is talking about oil. Oil discourse gushed onto the timeline this weekend. Crude prices spiked to nearly $120 a barrel as a broadening war in Iran threatens both transportation routes and production. The geopolitical and economic analyses are flowing. But what does this mean for AGI timelines? And a lot of people in the AI world are sort of tuning all of this out because they see recursive self improvement, AGI, asi, the build out as more important. And I just wanted to sort of reality check the AI supply chain to understand how does oil actually affect data center construction, AI production, token pricing? Like is there any effect? My conclusion was that it's very moderate, but there are some interesting effects in the financial markets that are probably the bigger takeaway. But it's still interesting to hear about. Like yes, oil actually does is used in the production of AI at least a little bit. So power has been at the forefront of the AI pushback. Like everyone's worried about these local energy prices, these electricity prices increasing near the data centers. But. And it's been, it's become like a political issue. But pain at the pump might become a bigger story as gasoline prices spike. And that has been pain at the pump. Oh, it's been the most tangible sign of inflation. It moves so quickly. You know, one jitter in the economy, it's a huge component. You know, people on the coast, people in tech don't have a good sense for this. Right. If gas, gas, for a lot of people, gas could quadruple and it wouldn't, they wouldn't really notice it. No, no. But if you actually look into the average American, how gasoline fits into their budget. It's a meaningful component of their monthly budget. So they feel it super intensely and variable. Yeah, cost. There are. There are so many different ways where a lot of Americans go on driving vacations that obviously is directly impacted by gas prices. And then also just psychologically, there's something about filling up at the gas tank where you see the number ticking up and you're doing that on a every week basis or so that it's just so visceral. It's this thing that you have to stop and then go experience and watch the money flow out of your account, like in real time. It's very visual. It's very interactive. Yeah, I remember. I remember I must have been probably 18 at this point where I would just go. And I would like for a long time. I just put, I'd like prepay for a certain amount. I got 20 bucks. I got 20 bucks. Let's see how much I got. And I felt like, really? I felt like the king of the castle. I just put my card down and let it run up. Yeah, yeah, yeah. No, no, Totally. So pain of the pump's been like a big, broad economic issue. It's so visceral. Like, the actual phrase has a lot of power in politics and in like attack ads. Like, your gas prices are going up under this person. That's why you got to vote for me. This is common, but in AI circles, the discussion's been much more focused on RSI now is the new acronym that everyone's focusing on. Not AGI. AGI is here. We know Artificial General Intelligence. We passed the Turing test. The models can do things generally intelligently, but can they recursively self improve? Are they rsi? Are we in RSI now? We. Is it coming? Is this going to be a fast takeoff? Is this going to be a slow takeoff? Well, something's taken off. Dylan Patel said being an SF is like being in Wuhan right before the pandemic. Something is happening. It's going to hit everywhere. But so few people know it. So he's sort of echoing that something big is happening mentality. This idea of like recursive self improvement being AGI pilled. A lot of smart people agree there's definitely something happening. There's still this question of, like, what is sticky? What? Where will diffusion take longer? What will be sort of AGI resistant or ASI resistant for a variety of reasons, even if they're completely irrational and you could get a better thing from a computer. But people like the Human version or whatever. The irony is that that George Hot's hitting the timeline to raise money makes me more bullish on acceleration. Yeah, because if not, he's obviously not historically been a huge fan of venture venture capital. Well, he's not raising from traditional VCs. He said he. He specifically said he wants to raise from like like basically like friends and family type of round. He does want 20 million. We have this in the timeline somewhere. Where is this post from George? I'll pull it up if you want to keep Tiny Corp. He said that he's raising $20 million to buy a building. He's. He's gonna buy a building. Here it is. If so he said this is from Tiny Corp. The makers of Tiny Grad and the Tiny Box. If Tiny Corp was raising $20 million at 200 million dollar valuation, who'd be interested? Business model is basically this. Buy an $11.5 million building with 5 megawatts of power. Link in our Discord, wait for AMD to launch the rDNA 596 gigabyte cards mid-2027 pre order 3000 cards. Hopefully we can negotiate for $2500 each. Build $520,000 tiny boxes with 6 of the cards in each box run all the Chinese LLMs make $600,000 per month revenue selling tokens on open router market depth is there. This is 1% of open router improvements to Tiny grad yield revenue improvements due to how power is priced in Oregon it's only like $50,000 for the electric bill before the 4 megawatts before they price for peak not usage we get like 3c kilowatt hour power $0.03 per kilowatt hour. We can also make $100,000 per month leasing colocation space to comma building and cards paid off in three years max investment made back low risk of being undercut since we're using consumer GPUs and running the cheapest colo you can believe. If someone chill wants in, I do it. I'm not going to fake hype. I'm not going to hype fake tech. But demand for tokens is going to skyrocket. Skyrocket. Look at the open claw install numbers. With crazy good optimizations we could potentially get 3x more from the machines and we have electricity 3 for 3x more machines, 5.4 revenue per month then continue to scale from there. Custom chips, et cetera. He's starting a Neo cloud or he's starting. Yeah, he's going to be serving tokens. So that was Incredibly bullish. Demand is definitely there. At the same time, he sort of took the other side of like, this is different. This is crazy. He says, there's no AGI, there's no magic threshold. You guys see auto research change the random seed from 42 to 137 and OMG, it's AGI, it's over. Yet you critique the junior engineer for the same stuff. The cost of development is falling. Overpaid engineering struggles to compete. That's the story. And he says humanity has been recursively self improving for centuries, so everything new is old. I don't know. I obviously have a huge amount of respect for George, but at the same time saying there is no artificial general intelligence when you're also arguing that engineers are struggling to compete with new technology that is replacing them. How is this not artificial intelligence competing with. Yeah, so I think he is agreeing with that. But it's that there is an immense amount of desire for this binary moment. This is the singular. This is AGI, this is asi, The RSI is here. This thing is happening right now. And there's before and after, and everything has changed. And he just doesn't see it that way. I think. I think he sees it much more like the Internet, the mobile phone, like other technologies that have been rolled out. Electricity. Yes, there is like a before and after, but you can only really define the period by maybe a decade. And you need a few decades to understand that moment. It's very hard to go back. And you know, there is like the iPhone moment and there is like, you know, the first launch of, I don't know, aol. Like, I don't even know. I don't even know what the iPhone moment of the Internet was. Just because it was sort of a slow rollout anyway. Live GPU clusters. Where's Jordy going? Oh, you're moving the goalposts. Okay, Jordy's moving the goalposts. Where are you moving the goalpost to? Just over here. Okay, but why? Is there a reason? Well, I mean, they have to be moved. George is moving the goalpost. Oh, okay, okay, okay. He's moving the goalposts. So I said, you know, it's clear that the AI industry continues to grow and continues to need more and more power and compute, as we've seen from George Hotz's new project. That means large data center campuses. But if they're not in random office buildings that George is picking up for on the cheap, they're probably going to be built with construction equipment. So what does this Mean they don't just drop from the heavens. They require building, which requires oil. But how much oil? And is oil a serious. And is a serious oil shock enough to impact the AI buildout in a meaningful way? Spoiler alert. Basically no. But live GPU clusters in the United States do not use much oil directly. Only 0.6% of U.S. electricity in 2024 came from petroleum. We're much, much more dependent on natural gas. Something like 42% of US electricity is natural gas. And so America ramped up natural gas production significantly over the past two decades. A lot of that was in reaction to the wars in the Middle East. Hey, we need to be less dependent on foreign oil. We need to be energy independent. And so you have the fracking boom, the natural gas boom, and that's where a lot of our fossil fuels come from. Today, I believe data centers only consume a single digit percent of US electricity. So you're looking at 0.6% of a few percent is like the actual impact. So the short term impact of high oil prices should be very limited. On AI building out new capacity is much more. That said, it's worth noting, I think it's something like Qatar produces something like 20% of the world's liquefied natural gas. Right. But not for us. Yeah, not for us. So we're in a good spot. But this is going to impact the data center build outs or active data centers across the rest of the world if this continues. Yeah. So when you're talking about building new capacity, building new data centers, oil is a little bit more involved. So diesel powers, trucks, trains, boats, barges, generators, pumps, compressors, excavators, and tons more construction equipment. Petroleum is also broadly used for plastics, polyurethane and solvents that all work their way into the data center supply chain. The biggest problem is delaying already tight schedules because of narrowly available components going out of stock. So the price of oil goes up. There's one marginal factory that can't produce one ingredient that goes into the rack and that slows things down. You have to wait a week while you find another supplier. That stuff can add up to just a little bit of a delay. This happened during COVID and the AI industry was already experiencing something similar with transformers. And so you don't want products getting stuck in transit or going out of stock. But the bigger problem and the one that people should be talking about, and I think you were debating with Dan Primack at Axios about this, is macroeconomics. So higher prices, higher oil prices lead to higher inflation. If the Fed has to raise rates to control inflation. Capital formation for mega projects gets a lot harder. So JL has this estimate. The next 100 gigawatts of data center capacity could require about $870 billion of new debt financing. And so using this rough number, every extra 50 basis points of borrowing cost on 870 billion is 4.35 billion in annual interest expense. And so that's, I don't know if that's a huge squeeze on cash flow, but it's certainly a squeeze on cash flow. If you're a hyperscaler and you're going gig along AI CapEx now and you're like, yeah, we're going to do 200 billion this year, we're going to do 500 billion next year or whatever across a couple companies. You start looking at interest expense bills that are already huge, but then they're going to go up by 5 billion or up by a full extra billion just with a 50 basis point increase, half a percent. So half a percent change in the fed funds rate can echo through. Even a quarter percent could probably be amplified into a multi billion dollar cost line on your balance sheet or on your income statement as you're servicing that debt. Overall, higher oil prices do make the AI rollout harder, but not really for existing AI capacity. I wouldn't expect token pricing to change based on the price of oil. But there is this more. There's this bigger question of macroeconomic resilience during a time where our largest tech companies are digging through the couch cushions to find every penny to win the AI future Middle Eastern investors pulling back. I agree with you. I do think that this is a significant risk because if they pull back on big investments makes these mega rounds more difficult. Retail investors could fly oil stocks or defense stocks. They think that this is like going to be a thing for months and months and months. It could all flip if there's a quick resolution though, which is what I'm hoping for. But if things continue, the AI industry walks an even tighter tightrope. Yeah, I mean the big question right now is the hundreds of billions of dollars of sovereign AI projects in the Middle east. Right. I think a lot of those people are going to be like, do we want to send billions of dollars of of GPUs over there? And then also the money coming here is another thing where you might want to spend it elsewhere quickly. Let me tell you about Turbo Puffer, Serverless Vector and full text search, both from first principles and object storage. Fast 10x cheaper and extremely scalable and Let me also tell you about public investing. For those who take it seriously. Stocks, options, bonds, crypto treasuries and more with great customer service, whatever investing strategy is. We have to say happy birthday. Happy birthday to Trey. Happy birthday, Trey. We love you. Excited to have you. Thank you. One of our strongest soldiers. We appreciate you. It's been an honor. Podcast. Have you been noticing that it's been hotter in Los Angeles? I have. Downtown Los Angeles is forecasted to approach 100 degrees Fahrenheit on Thursday and Friday in March, which is why not good. We're going to do the weather segment. We have a new segment for you today. We're doing the weather on tvpn. We have our very own Ben. We have been. Hello, guys. How are you? We're doing great. Tell us about the weather. What's happening? Well, I want to start off by saying, as you can see, the weather today for the low today is going to be 75 degrees Fahrenheit up there, high of 100 degrees Fahrenheit down there. But there's something I actually wanted to point out that I saw and I thought that was quite interesting. As you can see up here, there's a localized low pressure area up there and a localized high pressure area right down there, if you can see that. Right. Does that mean rain? The issue that's not normally an issue and not a cause for concern. And it's not very common for this time of the year. However, today, later in the afternoon, these two areas are going to collide and they're going to hit each other. No, really. And what that's going to cause is a barometric pressure inversion. Okay. It might sound a little bit scary, but I guarantee there's no car.