You're watching TVPN. Today is Monday, July 20th, 2026. Let's go to the fans zone. We are live from TV Ultra. The template technology, the fortress of finance, the capital of capital. Let me tell you about ramp.com. Time is money. Save both. He's to use corporate cards, bill pay, accounting, and a whole lot more all in one place. We had a fantastic talk with Eric, the co-CEO of RAMP on Thursday. But we are excited to be back. It's Monday and there's a lot of news. The big stuff, uh, Kimmy K3 dropped. But first, we should recap the weekend. How was your weekend?
The elephant in the room that I'm wearing a suit.
Yes, the elephant in the... Did somebody say that? The elephant in the room.
Jord's suit. He's feeling good. He's refreshed.
Yeah, refreshed. I uh, yeah, was in, was in the weekend... to the old country. Yes. Yes, exactly. Uh, really fun time. Glad to be back. Uh, very, very uh, refreshed, very fun. I was at a Caruso property this weekend and I got to say being there, being at the the Myiramar Rosewood, uh, I I can never vote for him for public office ever again.
Why? Because you want him to develop more properties like that.
I want him laser focused on resorts and malls.
It doesn't really hold up because the whole, the whole point was that he could, there was some possibility that he could make greater Los Angeles something like his incredible... got to get a Rosewood in every city in America and uh, maybe... or take private of the, of the city of Los Angeles, turn it all into a Caruso... potentially. But uh, it was, it was a good weekend. Glad to be back. Uh, the timeline was burning up over the weekend over uh, the AI cold war that's breaking out between the US frontier labs and Chinese open source. Uh, there's been a whole bunch of new Chinese open source models that have dropped with some uh, promising benchmarks. Uh, the latest one that everyone is up in arms about is uh, Moonshot's Kimmy K3. Uh, it was unveiled last Thursday. Now they aren't actually... Yeah,
I kept opening X and thinking, looking, scrolling a few times thinking, 'Oh, I'm happy for you.' Or sorry that happened. I put my phone away. I feel like I missed a lot.
Yeah, I mean it's, it's gated behind of uh, behind an app right now. Um, the weights are not actually open just yet. That should happen Monday if things go to plan. But of course there's a lot going back and forth uh, with the administration. People are uh, you know, uh, weighing every possible consequence of this. Um, it's a, you know, the classic, it's over, we're back. Is this good for semiconductors? Is this bad for semiconductors? Is this Jevons paradox? Is it the counter example to Jevons paradox? Um, it's all very interesting debates and everyone's uh, coming out with their own opinion. Dean Ball was going back and forth with uh, two people, uh, David Sacks and also uh, Emil Michael were duking it out on the timeline, fighting it out back and forth. Um, but you know, decent takes on both sides. Lots to chew through. Um, the headlines for Kimmy are impressive. Um, but not wildly surprising based on the current trend lines and model progress. Um, but it does seem to invalidate uh, the thesis that open source would start falling behind, which was something that was... there was that chart from the government I believe that showed uh, closed model progress was, you know, accelerating on this particular exponential and uh, open models were falling behind. Uh, this seems like a proper catch-up, something that was predicted from, you know, many lab leaders that, you know, the six-month gap, three-month gap, nine-month gap, it will remain just that gap um, for the foreseeable future. But there are of course a bunch of other nuances here. Um, in a, in the most above board scenario, sort of here's what happened. Like the Moonshot team, it's just a bunch of highly talented AI researchers who marshaled enough compute to release, to release a great model. Uh, open sourcing Kimmy is a great way to grab attention, attract talent, drive demand for hosted API services uh, that are more turnkey eventually. There's a very reasonable business model behind open sourcing. Uh, Red Hat Linux of course is an open-source company uh, owned by IBM. Now I think the acquisition price was in the $30 billion range. And at any point you could just say I'm just doing Red Hat on my own, but a lot of companies pay for uh, you know, consulting, hosted, and all sorts of different uh, services around the open-source piece of software. Uh, and so uh, that's sort of like the, the, the bull case, like, you know, everything's going well. The more skeptical folks on the timeline are pointing to uh, you know, ideas such as uh, maybe the compute was smuggled into China against chip controls. Maybe the data was exfiltrated from frontier lab APIs routed through wrapper companies. Maybe this is a distillation attack. Um, the open-source strategy is a deliberate attempt to destroy American companies. It's geopolitical warfare. It's the AI cold war. Um, and there might even be spies inside companies stealing IP directly. And so that's sort of like the bad version. Uh, the reality is probably a mix of both, but we'll know more over time. And we'll hear more statements from the labs and from all sorts of different uh, participants to understand there. People were going back and forth on, you know, if you ask Kimmy its name, will it just say it's Claude or GPT? And uh, there were some examples of people doing that. Seemed like it was photoshopped. Seemed like that. Also seems like the easiest thing to fix possibly. It's just put it in the, put in the prompt, put it in the fine tuning. Never can, you know, find and replace. Seems pretty easy.
They were showing it in the reasoning tracing.
Yeah. But, you know, I, I, I, I don't know. All these things are hard to, hard to assess over time. It takes time to actually work through what's real, what's fake, what's a hallucination. You know, they all, they all make mistakes from time to time. But um, there are bigger questions about AI safety in a world of powerful open-source AI. Uh, I haven't seen a ton of takes from this crowd, but you have to imagine that they aren't happy about it. Uh, maybe the uh, the stop AI protests will head over to Beijing next and beat down the, the doors of uh, Moonshot directly. Um, but it is an odd thing because even if you're not worried about the fast takeoff, you know, gray goo, paperclip, uh, from a cyber security perspective, like things are going to get more complex. There's reasonable solutions to all of this, but uh, it is a more powerful tool in the hackers or the, the hackers' tool chest. Uh, fortunately, defenders have had roughly six months of frontier exclusivity with GPT cyber and Mythos for a while and so there's been a lot of preparation. We've talked to Nikesh at Palo Alto Networks and George at CrowdStrike about this, like they have been aware that this is coming for a while and so a lot of systems have been patched. Uh, but at the same time you can just imagine now anyone who's a hacker can spin up a very, very intelligent model and start sending a bunch of spam emails or spam text messages that stop seeming so AI and they're not as clockable because it's just another level of intelligence and it can do uh, reasoning traces and research and, and so, you know, that the, the, the spammer or the phishing attack can actually know what insurance you have and, and pose as your insurance agent and actually know a little bit about your insurance agent instead of just sort of sending you a generic spam text. You're, things are going to get more customized, more personalizable. Of course, there will be more advanced defense layers, but the war is continuing. Um, there, there really is no debating that open source is just cool, like open source AI. It just feels punk rock to me. It feels like Napster, like as a kid, the idea of just like liberating all the information culturally aligned to the internet itself.
Yeah, it's just a cool thing even though you do have to grapple with all the different problems that come from it downstream. Uh, like the, there's something just deeply satisfying about the idea of uh, of like a solar powered server rack serving you up unstoppable intelligence, which is another good name for a neolab. Um, I don't know if there is one already but uh, you know, a country of geniuses. In some ways to me it feels like... aligned to America's ideals. Oh yeah. It's like... like it feels like, like very empowering.
It's total Second Amendment. Totally Second. Exactly. It's like I should be able to go off-grid and have uh, you know, not a country of geniuses in a data center controlled by a corporation, but just like my own genius who's riding with me no matter what in my basement, you know. And it's like, and, and you know, even, even if you're not thinking like nefarious, it's like this, it's this idea that, you know, uh, I like to have a copy of a movie in case like the grid goes down. You know, the, the more like, you know, apocalyptic thinking. This is a very general thing that I think a lot of people align to especially if they're at all crypto-native. It's like why do you want Bitcoin? Well it's unstoppable money. The government can't take it from you in certain scenarios. Of course like there's a bunch of ways they can of course like anything else. Um, but uh, there, there is just something very, very attractive about having just a machine with all of human knowledge on it. It's like a, it's like the, you know, every house used to have the, the full Encyclopedia Britannica and you can just look up anything and now you have the LLM that can solve, you know, any math problem for you. Um, but uh, you know, the US government is going back and forth on this issue uh, how to handle the situation and I imagine we'll see, you know, developments this week. Uh, the Kimmy K3 weights drop next Monday and so this is the week where if the government wants to do something one way or another um, they have the opportunity to do that. There was some back and forth on like would there be a, a direct ban, a hard ban, a soft ban, maybe just some like messaging around hey like please be a patriot, don't use this foreign product, you know, there's all sorts of different ways like the original like buy made in America, buy American. That's just marketing, but it does have some effect. It's not enough to make GoPro a, you know, 10 billion dollar company, but it is enough to, you know, cause a little bit of motion in the, in the, in the economy. Um,
One thing that was notable over the weekend, uh, Kimmy on X said that Kimmy K3 has received far more love than we expected and our GPUs are feeling it over the past 48 hours. Demand has pushed close to the limits of our current capacity to protect the experience of existing subscribers. We're temporarily pausing new subscriptions and prioritizing compute for current members.
We're adding capacity as fast as we can and we will reopen new subscription spots in batches. So again uh, Kimmy and the world more broadly are quite uh, supply constrained um, and uh, yeah, just showing or compute constrained and showing that uh, the other dimension of this whole geopolitical uh, battle is around the hardware side. So...
Yeah. Yeah. Um, I mean, in general, I think my, my takeaway was, and Tyler, I'd be interested to hear how you process this, but uh, like I, I just don't see this as a black pill, even if it's like a DeepSeek moment, the markets sort of bounce around for a little bit. Uh, it seems like there's still a lot of demand for frontier intelligence, a lot of demand for near frontier intelligence, a lot of demand for really cheap older models that have been baked down to basically intelligence too cheap to meter. And we are going to be in this process of go and find an actual use for something and then bake it down to be really, really cheap so that it can be offered by...
Yeah. And in so, so many moments over the last two years, we've, there's been a company that's come out, maybe they've raised some money, you maybe forget about them for like 6 months, and then they pop up and it turns out even though they're the number 10 company in their category, they're still growing faster than uh, almost any company uh, that existed like six years ago.
Yeah. And a lot of it goes down to deployment, too. There's like certain companies if they're in consumer like they just have a really good go-to market motion like they're just able to get customers to subscribe or use their product or if they're in enterprise like they just have a great like forward deployed motion to actually do the change necessary as opposed to just like here's an API law firm like they could have done that and yet you have legal AI companies that have done a great job of coming in and actually creating a product that works within the confines of the, the broader organization. Well, what, what are your Kimmy K3 uh, reactions, Tyler?
Uh, yeah, I mean I think it was definitely like a, a surprise, an update, like this is extremely good model. The idea that open source is falling behind seems like very much...
We sort of got a preview with GLM 5.2 like people were already saying like oh GLM 5.2 is sort of catching up. So...
Yes, but even, even that, that was closer to, you know, maybe a year, maybe a little less than, than a year behind. This seems like much closer.
But I, I think I, I broadly agree like, you know, we have, okay, there's this great open source model and now we have, okay, the Jacobian, you know, conjectures like this, this kind of storied math problem. Yeah, and so like, yeah, there's just going to be a lot more use like everywhere. I, I expect, you know, these kinds of models to, to be used everywhere.
It seems like a lot of the hype is around like specifically the parameters, the size of the model, the way it was trained. It'll be, and then also it did very well in the front-end arena which was interesting because that's, I mean it jumped ahead of everything there and that feels like a particular like, it's, it'll be interesting to see because you see more and more taste going into these models like OpenAI is really big into the math stuff and uh, Anthropic was really big on code early and there's all these different like, you can just tell that like oh this lab cares about chess now and now the model's good at chess or the model got better at writing because like they, they took the time to spend a month sprinting on that and it'll be interesting to see like where did they sprint if anywhere, where didn't they sprint and it's sort of lagging uh, and then what harness it works best with, how, how, how important is the harness relative to everything else but uh, we'll be, we'll be interesting to, to continue following. Um, there's a whole bunch of great reactions, SemiAnalysis uh, it breaks down the architecture. Uh, where should we start here? Let me tell you about Cisco first. So critical infrastructure for the AI era unlocks seamless, less real-time experiences and new value with Cisco. Um, there's a, there's a long post by...
Kimmy achieved another uh, pretty meaningful benchmark.
Uh, oh yes, on the Belgian Grand Prix.
Is there anything these Chinese models can't do? Says Alex Conrad.
Kimmy Antonelli of course of the, of F1. Uh, anyways, over to SemiAnalysis. What do they have to say? Where, where are they here? Uh, SemiAnalysis said similar to the panic over DeepSeek R1, some uneducated people think Kimmy K3's use of linear attention KDA is bad for Nvidia HBM DRAM and networking because it has relatively lower KV cache requirements. The opposite is true and they'll explain. Kimmy K3 is actually quite positive for Nvidia as large model inference is where the NVL72 shines because K3 has more than 2.88 trillion parameters. It requires a large scale-up domain to store its weights. Secondly, although Kimmy delta attention has up to 10x lower networking requirements for KV cache transfers, its large weights require even more network bandwidth to implement an optimization called wide EP which spreads weights across different GPUs. It is interesting that uh, I mean it feels like if, like is China getting TPUs? Those have to be uh, chip export restricted and probably more difficult to get into the country because there's less resellers and in order to set up a TPU based data center you probably have to work very closely with Google whereas Nvidia GPUs are sort of going to like all sorts of small Neoclouds and there's a lot of places where seven steps down the chain it can wind up in a suitcase with a, with a hair dryer removing the label or something like that.
Well, it's not even that. You can imagine like if you were a Chinese lab and you wanted to get access to compute, there would be a number of ways to like set up US entities that then basically get compute through variety of different sources here in the US. Yeah.
And you're, you're uh, just getting it remotely. Yeah. Uh, CK Capital says, 'The hottest model in the world just started turning away paying customers because there isn't enough compute to serve them. This is the Neocloud thesis in one uh, post.' Um, and uh, the, the Neoclouds were sort of getting beat up, was it last week or the week before? And uh, and it seems like they should be back. I don't exactly know what uh, what's CoreWeave...
Yeah, CoreWeave is up uh, 21% today. W... I mean CoreWeave only... they've been, they've been sort of a, a laggard and down over the last 5 days and the last month. Um, but yeah uh, if you're, if you're flexible and ready to serve this the day it comes online that's probably profitable very, very quickly even at, you know, lower than expected margins. Um, uh, Nicholas Bamante at Microsoft uh, who is working on AI for knowledge workers uh, says, 'Guys, the world is so compute. I honestly don't understand how anyone thinks otherwise. Almost nobody is seriously using AI today, and we're already hitting capacity. Everything is a bottleneck. Land for data centers, permits, electricity, grid connections, chips, construction, cooling. Demand is growing far faster than supply and expanding supply is insanely expensive in this game. The best positioned companies are the ones with massive net income from non-AI products. They can fund hundreds of billions in capex.' He's given a little bit of a bull case for Microsoft. Hundreds of billions of...
Mo over at Vercel says...
Oh, yeah. That's good to hear.
Yeah. Kimmy K3 is top tier at cyber security.
Uh, that will uh, that will be interesting as the, as basically uh, the weights are released. Do we see an uptick in uh, attacks?
And it's also sort of a, a reputation of like the, it's just distilled because the models that you would be distilling from are pretty severely nerfed on cyber security at least in theory. And so it's like you would need to distill on the non, uh, the non-nerfed model which is probably way harder because those are much more restricted. And so uh, it, it, it does feel like there's a uh, uh, there's a little bit of a reputation of the idea that these are pure, purely distilled.
Uh, Soul is a leap ahead in cyber capability at a significantly higher cost but quite remarkable still. Fable refuses everything. We couldn't get it to complete the run at all. What is interesting is that Soul in comparison was much more open to helping with defensive cyber hardening. Uh, TLDR: frontier open weight cyber capability is here and um, yeah, they have uh, their own product for uh, defensive cyber security over at Purcell. Um, this was good. Hi Harry says rip the bear case July 16th to July 19th perfectly coinciding with my uh, trip uh, which, which is, which is nice.
Well let me tell you about Console. Console builds AI agents that automate 70% of IT, HR, and finance support, giving employees instant resolution for access requests and password resets. Um, Dean Ball was going back and forth. Uh, oh, some deleted posts now. Um, but uh, his, his observations were getting taken to task going back and forth. He says it's a very good model. I don't think its performance can be explained away by distillation or anything like that in agentic coding sessions. It seems pretty much on par with the best public models of Q1 2026 in my fairly limited use. It also seems very token hungry. There was some interesting back and forth around the pricing because it might not actually be that cheap at least through Kimmy's API or Moonshot's API. But...
Yeah, I think the, the input per million was like $3, output was 15, which is like uh, cheaper than frontier models, but like it's, it's certainly not at like...
Yeah. And if, if the, if it's like more token hungry...
Yeah, token efficiency is not very good. That's actually going to be much...
But you have to imagine that if it's open weights, then, then companies bring it in, they start quantizing it and distilling it and tightening it up and optimizing it and then there's probably a whole stack of like Neoclouds that will be competing for, well, you know, we have this cheap energy over here or we have some older chips that it can run on or do whatever they can to, to bring down the price.
Uh, Dean says, 'I am personally surprised the Chinese state continues to allow the open sourcing of models this good given potential risks.' Now, it seems like the, that is very much like from the top they are pro-open source right now. Um, Dean says, 'To be clear, I myself might be fine with models presenting this level of marginal risk being open weight, but I'm surprised that China is fine with it. I suspect the reason is that they uh, they are is 75% explained by strategic blindness, lack of AGI pledness. The CCP is very lean lacuni in its view of AI. Uh, the other 25% or so is their lack of compute for customer inference making China's open weight strategy an unintended product of US export controls.' That's interesting. Uh, and the normal Chinese strategy of aggressive exports for the companies as opposed to the government. The decision to open source is partially ideological and partially because they are behind and they know that very few people would pay for subfrontier models from China. Yeah, you have to imagine that it, like the idea of sending, I mean people are having discussion over like should you send your corporate data to a, to an American frontier lab. Imagine that that same level of should we upload our entire codebase to the Moonshot API because we've gone all in on, on Chinese closed source which seems like that's a very, very hard sell. So there is a world where there's like no money to be made in America. I don't know. Maybe there's some American businesses that would just be like yeah I'm cool with it. Like take my, take everything, take all my IP. Like I, I'm, I'm, I'm uh, uh, what's the phrase? Uh, like anti-fragile. I'm anti-fragile. I can't be like, you can't clone me. So, and I have no IP, so just upload it all.
I mean, if you're a local plumbing company.
Maybe. Maybe. Uh, he, this is where he gets uh, pretty, this is a pretty hot take. Dean says, 'Open weight models are inherently decelerationist.' And Tyler and I were debating this before the show. Uh, he said, 'I'm continually surprised to see the so-called accelerationists so excited about open weight models. I suspect the reason is that they know open weight models are effectively ungovernable.' That's correct. Uh, and they simply like the overall cloak of ungovernability open weight models uh, create over the whole of AI. It's not a bad strategy. It reminds me of James Scott's recounting of the hill people in The Art of Not Being Governed. Still, in the end, open weight models deter further AI capex. And this is the debate that we were having was like how capex intensive are these frontier training runs because Moonshot's raised I think a billion dollars recently uh, and, and it just feels like even them, even the latest labs like there's a lot of cost, a lot of researchers and they do a lot of experiments but the, but the core run might not, we might not be talking about 50 billion yet maybe in the future like you, you need to accrue towards that and maybe people wouldn't fund it if they're like ah you're only going to be able to monetize this for, for like three months because you're going to get distilled so fast or you're going to get copied so fast. That is tricky. Like it is nicer when you have a monopoly or oligopoly for some period of time to, to actually earn profits off of what you invested against.
Well, if anyone listening to this is over at Kimmy or Moonshot...
No, don't come on the show. Just uh, feel free to let us know how, how you made it.
Yeah, maybe. Feel free to, to leak the secret sauce...
On the show. Come on the show.
Yeah. Yeah. If you're so, if you're so open, why not? Uh, so growing Daniel was, was calling this dumping. He said what China is doing in AI is called dumping. They do it in literally every industry they enter. The goal is to kill all local competition by subsidizing their own industry so they can produce at a loss. Then once all competitors are dead, they can charge profitable prices and control the market. In steel and automobiles, this is just as, just bad. In AI it's potentially fatal, fatal to our country. Um, it's interesting, it's like I, the, there's a deeper AI supply chain across data centers and fabs and chips and energy that you could potentially have a lot of GDP growth and a lot of industrial AI capacity on an open weight model. Um, so it's not like you're losing everything, but it does feel like a little bit of a slippery slope where you're like, 'Sure, like we'll, if you let China win the, the, the model layer, then you also are like, sure, they set up a fab like let's start buying the chips for them. Sure, like just put the data centers there. Put everything there.' And then all of a sudden you've lost like everything potentially and you're just in that consumer mindset and then you're back to well, how hard is it to make a t-shirt in this country? Is that a problem? Uh, not if we're friendly and we're trading and we have something to offer, but if we don't have anything to offer, then we can't buy anything and then we become a poor country. I don't know. Uh, it's a, it's a tricky back and forth. People like cheap stuff, like they like cheap TVs and they don't like pollution associated with TV factories, so you send them abroad and then eventually you're like, wait, what do we make in this country? Uh, anyway, we have Tyler Cowen joining at noon. We can break down all the economics of this and other topics in just a minute. Uh, let me tell you about the New York Stock Exchange. Want to change the world? Raise capital at the New York Stock Exchange. Just do it. Uh, growing Daniel says, 'Unfortunately dumping uh, unfortunately dumping is the very common argument for rent seeking domestic firms who want protection. So even if you're in just a normal competitive industry, uh, if you're an electric vehicle manufacturer, you're going to claim dumping even if it's just fair competition from BYD. Uh, even if there's no government subsidies, they're just making great products and they want to send them here.' Um, Dean's position here seems to be a Jones Act of sorts and this has famously not saved our ship building industry and I suspect it won't save our AI labs. Stopping open weights is virtually impossible. So our only other option is governments taking stakes in labs and subsidizing our own industry. Uh, if you scream and yell about either Jones acting, Jones acting AI or subsidizing GPUs, then you should also explain how we will escape the dumping trap China is trying to push on us to destroy our labs and leave themselves with the only functioning AI industry in the world. That is a take. Uh, Dean Ball says, 'I know I've said this a bunch, but I wasn't proposing a Jones Act for AI as a good thing. I was saying that if USG feels pressure to do something, the likeliest thing they'll land on would be soft law discouragement of Chinese model use. China does the same to our tech.' Um, yeah, that is true. There's a whole bunch of, I mean, even with the Nvidia chips, for a while there was rumors that uh, Beijing was recommending that uh, tech companies in China not buy them even once they were like approved. And so uh, soft power of the government happens uh, all over the place. Uh, uh, LeBron chimed in. Augustine LeBron that is, uh, former guest of Dwaresh Patel. I really like uh, him. Uh, he says there is no predatory dumping trap to escape. Keep thinking it through. One, China subsidizes open weight models below their cost, taking big losses in the process. Two, say they put Anthropic and OpenAI out of business. Three, they start charging monopolist prices to recoup losses. So, uh, now there's room for new entrants and perhaps government support. The monopolist prices era will be short-lived. All these complaints about unfair competition are BS. If you don't want to be serfs to a more competitive rentier class, then go compete. The more I think about the China dumping open weight models at a loss to hurt the US idea, the stupider it sounds. Now, you can lose the industrial capacity over a generation. Like if you, if you truly like lose all your AI researchers and they go off and do biology research or they, they just go do something else. And this is what happened. Like we just don't have that many nuclear engineers in America anymore because nuclear engineering was seen as like a dead end. And so they all went into computer science and now we're really strong in computer science and that's great. But if you, if you lose all of those people eventually you can't just like flip the switch and recompete. But I, I take his point that it is difficult to uh, to, to like you do create a market incentive. It's like wait there are Chinese companies that are charging us 99% margins on intelligence like any venture capitalists want to fund something a neolab perhaps and people will just be like absolutely I would love to compete with China like people do this all the time. Um...
Yeah, we're doing this in, in manufacturing.
No, yeah, yeah, we're literally doing it right now. It is slow.
Yeah. The automotive example is interesting. Um, in uh, over the last few days I saw a ton of BYD cars.
Uh, just all over. Uh, and it's interesting because uh, they were, I saw, I, I would say like less than I would have expected based on what a lot of people say on X, for example. They're like, I was in Europe and there's...
Chinese cars everywhere. I saw, I saw plenty. But what's really hurt uh, the European auto manufacturers is not that the Chinese companies are selling into Europe. It's that Chinese consumers are just buying Chinese cars, right? And so you look at a car like the uh, like the Porsche Cayenne, right? That was a big seller.
Huge seller in China probably still, they still decent amount but uh, people are very frequently uh, choosing local options and they have a lot of national pride in China around buying uh, local at this point.
Yeah, I like this final post from Augustine LeBron. Uh, he says the SFAI scene needs fewer philosopher kings and more CEOs laser focused on serving the best possible tokens at the best possible price. Your margin is my opportunity. Less Marcus Aurelius, more Ron Vakris. Costco CEO started his career as a forklift driver. Can you imagine if in a decade we have a uh, we have a, a frontier lab CEO who's like, 'Yeah, I started digging ditches to build data centers.' Like, I was racking GPUs when they were fighting it out over who was on the frontier and now I got the cheapest models in the business. I got the cheapest intelligence.
I got the best value in the industry. I mean like for, for like 99% of businesses, they just want tokens that are good at the right price. Like um, but maybe, maybe if Costco is the model, there will be a, an LVMH of AI. Something that really says your, your, your company takes luxury tokens seriously. It's like I'm, I'm happy to pay 99% margins. I don't even think...
These aren't just tokens. Yes, they're a statement.
This is a statement. At this company we, we use premium AI, luxury AI. Uh, Jakan is sharing uh, some news. The US is tougher curbs on Chinese AI. Now there are like logical curbs, right? Like distillation attacks and IP theft should be investigated. And I don't know what resources there are to fight that but I, I mean I like the FBI busts people all the time. Uh, there was a, there was a spy at Tesla a couple years ago who was like just straight up like saving the files, sending them to China, you know, like that happens and they bust those people. Um, and so there should be like an effort there and I'm sure there already is one. Um, but uh, Jukan says, 'Damn, this is exactly what Dean Ball said, isn't it?' And here's the news from Axios. The big picture, the administration wouldn't need to impose an outright ban to get US companies to drop those Chinese platforms. Quote, 'What's actually happening is slower and more durable.' Uh, one source familiar with the government decisions said, citing procurement rules, entityless threats and public pressure campaigns aimed at US companies using Chinese models. Instead of a ban, another source familiar with government discussions described a push to highlight potential backdoors and lack of security with Chinese models. It feels like it should be pretty doable to hunt down backdoors and deal with security risks of Chinese models. Um, especially if they're being run in American data centers uh, and the governance issue that brings uh, it's an offensive approach where the administration encourages a more innovative US open-source ecosystem the source added. So, uh, this will be playing out over...
Says token atelier.
Token atelier. I like that.
We'll take your token atelier over to Codex. It's a powerful workspace for getting work done with AI agents. Whether you're writing code, analyzing data, creating content, or automating business workflows, Codex helps you move projects forward from start to finish.
The Odyssey is up to 264 million. Biggest non-animated opening of the year until Spider-Man in two weeks according to Lucas Shaw over at Bloomberg. Um,
For reference, $375 million budget. Production was $250 million. They recouped that uh, immediately. $125 million in global marketing apparently. Um, and looks like a big financial win, but very much, you know, priced in like uh, you know, the base case here. Uh, the film will make a little bit over a billion dollars. Uh, the cash flow to Universal and Comcast, who are uh, the studio behind this, uh, you're looking at $250 million in pre-tax lifetime profit. That's 2% of Comcast's market value. So, it's not like this crazy thing for the corporate entity, but it is exciting for the film and it's exciting for Christopher Nolan, who's in the journal today in the business and finance section uh, because Nolan's The Odyssey shows directors are franchises. Now, they have an interesting statistic in here that uh, 53% of Odyssey attendees said that the director was their number one reason for attending. Not the actor, not the fact that they love the book and they were waiting for the IP to get adapted. Not that it's a, you know, oh, they like Sword and Sandals, they were looking or that it was IMAX, they wanted a Nolan film. And that's, I mean, everyone knows this intuitively from talking to everyone. Everyone's like, I want to see the next Nolan film. But uh, it's just interesting that, that there has been this huge shift towards uh, franchises built around particular intellectual property verticals, Marvel, Star Wars, Disney stuff, Pixar and whatnot. But uh, now people are saying I'm, I'm just going to ride with this director and whatever he does and he can take me to the past, he can take me to the future. Wherever he goes, I'm riding with Nolan. I had a funny run-in with Christopher Nolan. I was uh, walking on the beach and uh, Christopher Nolan was walking in front of me.
I'm not hitting any sound effects right now to be clear. Um, and uh, he's walking with his dog. I don't know who he is or what he looks like. Someone else who I'm with says that's Christopher Nolan. I said, 'Who?' Uh, and then, and then uh, it came to me. Uh, but then this dog ran up behind us. He was walking his dog. Another dog joins and is like really messing with his dog. And so I had to be like, I just had to say, 'Hey man, it's not my dog' because it seemed, it seemed like my dog was off-ramp. I was being framed for having a poorly trained dog that was harassing his dog.' Um, but he, he was pretty unbothered. Um, let's head over to this picture from Goier. He had the best seat in the house.
This is, is this a real photo or do you think this is like take a photo and then put it in AI and say like make it even more distorted because this is what it looks like to be truly in the front row of an IMAX uh, showing. Uh, people really were really crazy about like the IMAX looked way better. I think some folks on the team saw the film once, twice, three times. Uh, who saw it in IMAX? Okay, we got some IMAX chat back yet. Mike's seen it.
Give us the review. You liked it. What, what, how many thumbs up out of 10? What, what do you got for me?
Tyler, you can, you can chime in, too. What do you think?
Were you taking, were you taking notes the whole time? Like, did you have a, a notebook?
Uh, no. Mentally? Mentally? No. But it, it, it's great. It's an amazing accomplishment. You're like, 'Shoot the next movie in IMAX, dude.' I mean, it's one of the best looking films I think ever. I mean, the quality is just like unmatched.
70 mil IMAX. It's, it's, it's incredible. It's incredible.
How many times do you think, how many times do you think you'll see it in theaters?
Well, I mean, I'm already scheduled for four total.
Four total? That's crazy.
We'll see. We'll see how that go. I was slumming it in normal theaters. I, I don't want to spoil it in case you haven't read it yet.
Okay. Yeah. Thank you. You know.
Just, yeah. But I, I think uh, my...
I'm really excited to find out if he gets home.
Uh, my, my take I posted it was missing some, some high jinks and some gas. There, there are a few scenes missing from the actual poem that I, I would have liked to see. I think it was...
Wait, are there high jinks in the poem that you didn't, that didn't make it in?
Yes, that's what I'm saying.
It was, it was low on high jinks.
We should do an episode on IMAX.
Oh, yeah. We got to film this show in IMAX. That'd be awesome. Um, how would you rank it in the Nolan canon? Top five, top two. What are we doing?
I, I thought it was, I don't know. Production team. It's not top three for sure.
Not top three. What are you putting above it?
He's not going to beat the Dark Knight.
Yeah, Dark Knight. Dark Knight's better.
Inception. Tenet. Obviously,
I think it says a lot about the filmmaker. If, if this film doesn't make his top three for some people that says a lot about, or top five just with like his filmography.
Top five right now. Top five Nolan films go.
Inception, Dark Knight, Interstellar,
Oppenheimer. Odyssey's got to be up there.
Okay. Odyssey is in the, in the top five.
There you go. Well, hopefully uh, Jordy can take a look at it. I think we...
Said we'll wait till I can chime in to solidify that top five. Let me tell you about Public Investing for those who take it seriously. They got stocks, options, bonds, crypto, treasuries, and more with great customer service. Uh, yeah, the uh, the Wall Street Journal has this article, the no uh, the Odyssey proves Hollywood top directors are franchises now. Uh, every Hollywood studio wants its own Christopher Nolan. And I'm interested to know like how locked in is he with Comcast particularly like Universal? Like it, it feels like it's this huge asset, but it's wildly different than, you know, when you're Disney and you just own Marvel and there's absolutely no key man risk. Um, but it's clearly worked out very well. Uh, The Odyssey was written and directed by Christopher Nolan. It opened to an estimated 600 uh, 264 million worldwide domestically. It's 124 million debut was the third biggest for any film this year. So, it's not that huge. It didn't beat Toy Story 5, didn't beat Super Mario Galaxy, but it does feel like it'll have a longer run of people seeing it multiple times, paying...
More for IMAX, and it is just a breakout for an R-rated historical movie. That's a niche that typically doesn't put up massive numbers, like a Toy Story, for example. The opening was higher than many 2026 releases based on current, more current intellectual properties. It smoked Star Wars, The Mandalorian, and Grou. It smoked Supergirl, and it also beat Scream 7. Nolan's brand power and commercial track record got Comcast Universal Pictures to say yes to a roughly $250 million R-rated adaptation of an ancient Greek poem shot across Europe with largely practical visual effects, including a 35-foot tall Trojan horse. According to data from Renrak, 53% of the Odyssey attendees said the director was the number one reason for attending. 'It's not often you see that,' said Universal president of domestic distribution. 'Chris Nolan has earned the audience's trust and respect.' And that's completely true. Tenet had a very, very weak box office opening. Oppenheimer did quite well. Dunkirk did well. I recently rewatched Dunkirk. That's a great one. That needs to be up there in the top 20. They're all good.
In the late 2000s and 2010, franchises dominated the movie industry. Studios hired directors who could pump out sequels, spin-offs, and remakes as efficiently as possible. Nolan was the most successful and prolific among a tiny number of filmmakers who still called their own shots. Along with the hugely successful Dark Knight trilogy, he made his original hit movies like Inception and Interstellar. But Marvel, Fast and Furious, and Transformers have all stumbled at the box office recently as audiences have grown tired of them. Gen Z, in particular, appears skeptical of the industry's reliance on aging brands and technology-driven filmmaking. 'They're more drawn to stories told by an authentic, recognizable filmmaker with a point of view,' says the Wall Street Journal. Like Curry Barker, director of surprise horror blockbuster Obsession. But that film was one of the several new to the big screen director-driven hits this year, including Project Hail Mary and Back Rooms. In response, studios that used to obsess over franchise management are devoting more energy to signing the most promising filmmakers. So, you can imagine a 10-movie deal. I think Curry Barker already has a multiple movie deal just on the back of Back Rooms, even though it's his first film because it was so successful. 'Studios are lost as far as knowing what works and what doesn't,' said Byron Lord, CEO of CIA. 'Things have generally generationally changed and individual singular voices are taking advantage.' Warner Brothers gave Black Panther and Creed director Ryan Coogler an unusually generous deal to the rights to make his own original horror movie Sinners, which became a hit last year. Weapons director Zack Kger is helming an adaptation of the video game Resident Evil for Sony Pictures this fall and was part of a recent multi-million dollar deal Warner made for a new horror property. Barbie director Greta Gerwig used her clout to convince Netflix to make her adaptation of The Chronicles of Narnia.
Interesting. They're rebooting that. Resident Evil has had a rough go. They've done a number of movies but never a really major breakout beyond the core fan base. Was there a movie that paired? We were so close to getting another Barbenheimer with the Odyssey. Was there another movie that people could go see at the same time for the double feature?
They got to do that. No, they kind of... it was the Citizen one, Tyler. No, just kidding. I don't know. It would have been good. It made the whole Oppenheimer weekend so much more enjoyable having like the contrast there. Nolan should have almost... almost...
Did you do the Barbenheimer same day? I don't think I did it same day.
Wow. Same weekend though.
Laring over here. But it was fun. It was a good movie.
Netflix disclosed it has used generative AI in about 300 different productions this year already.
People are going to love this.
They wrote across the production life cycle from concept and previsualization through post and delivery.
'GenAI utilization by our creative partners is scaling quickly. In 2026, GenAI workflows have been used in roughly 300 of our titles with the largest concentration of work in post-production. We are increasingly leveraging these tools to deliver higher quality output more quickly and at a lower cost than traditional methods.'
Yeah. One of the clearest examples came from the American Experiment, a docu-series about the American Revolution. Co-CEO Ted Sarando said the series included 17 minutes of AI-enhanced footage that explained the scale of the project and that expanded the scale of the project and would not have been financially feasible using traditional production methods. The company insists that AI is not being introduced to replace writers, directors, actors or other creative professionals. Instead, Netflix is positioning the technology as the next evolution of filmmaking software, allowing productions to create shots and sequences that would previously have been impossible or overly expensive. Very cool. I don't know. I think if used tastefully, it can be good. I saw the designer behind Quarter, those quarterly earnings posters, put together a number of Odyssey film posters and immediately got community noted because his post was like, 'I made, I went and saw the movie and then I made these posters,' and a bunch of people liked them, and immediately got community noted he used Midjourney and, you know, he's just...
All right, here's a question. Is this AI?
Honda Canada posted an ad for the Odyssey starring your son, your son's BFF, you, your daughter, backseat snacks, and your dog.
It's a good ad. Might be, could be AI.
Could just be Photoshop? I mean, they just have a picture of a...
Well, no, no. I was like genuinely curious. Is this a... Did Honda actually post this? Like, you just never know at this point.
Oh, yeah. Yeah. I think they did. I think because it's Honda Canada, it's a more regional account. They can be a little bit more willing.
I don't believe that it's real. Plus, Trunk Fan, he knows when to flag things with real or AI or fake. They say the cast of characters that have been in the backseat of your family's Odyssey could make a movie of their own.
Yeah, man. The last drive in in your Honda Odyssey was a motion picture.
This is crazy and disgusting and disappointing.
Apparently trial lawyers are lobbying against self-driving cars because they're too safe. They need people to be killed and injured so that they have material for lawsuits. It says so with thousands of lives annually in the balance, who is against autonomous vehicles? Trial lawyers. Remarkably, the trial lawyers saw the writing on the wall very early and have been lobbying against AVs for nearly a decade. The American Association for Justice, the trial lawyers lobby, has been a prominent opponent of AV legislation.
It's crazy that on my drive to work, at least 80% of all billboards that I see are for personal lawyers.
Yeah. The Morgan and Morgan founder is a billionaire off of it. I mean, he scaled that business. And I saw some viral clip of a guy who just does lead gen and makes 30 million a year.
And doesn't actually, he needs to be a lawyer to put up the ads, but doesn't fight the cases, just finds the client, finds the plaintiffs, vets them, and then passes along to another lawyer who actually will fight the case, get the settlement. And so there's a little bit of back and forth about this. It might have just been that they're lobbying to hold the automaker liable so there's still someone to sue and that might be a little less egregious. But if that's what delays the rollout of self-driving cars, that would truly be disappointing. Nick Carter says, 'My jaw dropped reading this.' Who is the most opposed to self-driving cars? Ambulance chasing lawyers who sue over car accidents. They are explicitly on the side of preventable death. Not good. Oh, that's really, there's some really dark potential secondary effects. It fully rage-baited Alex Tabarrok, who wrote about it at Marginal Revolution, and we can talk about it with Tyler Cowen in just a few minutes when he joins. There were people going back and forth over Brian Chesky. Was his account hacked? Was it not? There was an AI-generated quote slop thread about crypto or something like that.
I think it was about, yeah, real world assets.
Which could be something that he gets excited about. It does feel like a little bit off topic right now in a world that is so AI-dominated, no one really wants to...
Yeah. He's in the mid, like the rumor...
There's been reporting that he's in the midst of raising for a Neol design. Well, yeah. Why would he be posting about that?
And so he told Fortune that his account was hacked and a lot of people weren't buying it. But Max Sparrow, the king of slop detection, the slop janitor over at Pangram, says, 'I keep seeing people skeptical that Brian Chesky's AI slop crypto thread was not written by him, but to me it seems like the most likely story. My guess is that the hackers were trying to warm up his account before dropping a pump and dump coin, but the actual scam post got flagged before it ever went live.' And so that's sort of an interesting twist because people were saying like, well, he might have just decided to post it because he thought it was good content or whatever, could get some virality. And the fact that it didn't have an address or wasn't immediately like the obvious scam, that was evidence that it was in fact just him slopping it up on the timeline. But this warm-up thesis does seem more likely to me.
Yeah. And if you had access to Chesky's account, you probably would think like, okay, do the warm-up post and then do something related to like real estate on-chain that is monetized.
Yeah. You got to bridge them. You can't just immediately be like, 'I'm giving everyone Bitcoin. Like send me your address. Like send me one Bitcoin, I'll send you back two.' Like that. People that probably don't even fall for those scams anymore. Hopefully. I don't know. People were going back and forth on this. It was a fun time on the internet. What else is going on? The Jacobian conjecture was solved. Another math problem. False.
Look, if this is surprising to you, if you needed this explained, or you're surprised that it fell, you're telling on yourself.
Yeah. Yeah. You are telling it to yourself. I like this from Bonito Torrellini. They just created a million Jacobian conjectures. You have no idea what's coming. But no, it's a very exciting another math problem. Also an interesting full circle moment because the mathematician who solved the Jacobian conjecture had been working on it, I believe, for like a decade. So he's an AI researcher at Anthropic now. And he, you know, was able to use Fable to, you know, prove this math problem that he had been struggling with for a very long time. Was it him or was it his advisor? Like he had worked on this for a while or been exposed to it at least.
Yeah, I think loose, like same area of math also. He didn't prove it. He a counterexample. Counterexample. Yeah.
Okay. Well, for everyone who's like, 'Oh, do I really need these data centers? Like I don't like AI.' Like this should swing them. Like put this on a billboard. Like I got to send this to Sager and Jetty and let him know like, 'Hey, this is, there's populism is going to swing around on this now.' I think...
Was this conjecture on Sagerbet?
I don't think he was. We got to get it on there.
Oh, he's kind of asleep at the wheel if he's, if he wasn't, you know, trying to make a...
Now he's in charge of this thing.
No, I mean, you know, on Sagerbet, he should have had, he should have had a market for this.
For sure. What else is going on in the timeline right now? Five cups of coffee. You got to be drinking five cups of coffee. This is the new meta. This is what everyone recommends. The Wall Street Journal has an article here. Got to pull it up. Where is it? WSJ. Five cups of coffee. How many cups of coffee do you do per day? I do one cup of coffee, but then I do two Aeromatics and three Diet Cokes, which I think is around five cups of coffee in terms of caffeine.
How much, how much caffeine is in a cup of coffee?
I do a Celsius on the way to the gym.
Oh, you drink caffeine before the gym? That's smart.
Oh, yeah. I have to. I have to.
Celsius on the way to the gym.
Regular coffee during breakfast. And then two of these.
Okay. And then you're done for the day.
Then I'm done for the day.
Okay. Okay. Five cups of coffee a day is fine for most adults, the Heart Association says. Tyler, did you get me? What's in a cup of coffee?
95. So, so putting up 500 milligrams a day.
So, that's like two Celsius. Two, maybe one and a half depending on which, how strong it is. They're between 200 and 300, right? Celsius.
I think they're up there.
200 is the max. I think...
No, I think 200 is the min.
I thought, I thought, I thought 200.
Yeah. So, beverage companies start to get...
Performance Edition and it goes up to 300 because they have a bigger can.
And they have Celsius Heat, which is 300.
Spice up your mind. Anyway, research has been mixed but association says evidence of coffee's safety and even benefits has been growing over time. This from the Wall Street Journal. This is important for everyone listening. There's good news for coffee drinkers. Good news for coffee drinkers. Up to five cups of the brew a day are safe for most adults and may even have benefits for heart health. That is crazy to hear. It always seemed like it would be a Faustian bargain at the very least, like, you know, worth it in the short term but at some point it's going to come back to bite you, but not in this case.
I mean, like famously Voltaire would drink like 40 cups of coffee a day.
Maybe we even know this stuff. That's...
You did pretty well. I would say if you learned that Voltaire drank 40 cups of coffee a day from Tyler right now, you're telling on yourself.
You're definitely telling on yourself. Well, let me tell you about Railway. Railway is the all-in-one intelligent cloud provider. Use your favorite agent to deploy web apps, servers, databases, and more. Well, Railway automatically takes care of scaling, monitoring, and security. And without further ado, our first guest of the show is in the waiting room. We have Tyler Cowen, the American economist, to talk to us about all things. How are you doing, Tyler? Good to see you again.
Hello. How are you? I don't drink coffee. Anyway, I have to ask every guest. Are you going to start?
I've only had coffee twice in my life. Both times in rural Ethiopia in a coffee ceremony and it was amazing, but I thought this is the peak. I can't beat this.
Is that because you'd become too powerful or do you see it as a Faustian bargain? I see people getting addicted to it and if they can't get their morning coffee, they have headaches. They don't want to be there.
Yeah, it does become a supply chain though.
How often are you somewhere where you can't possibly get a coffee in the morning?
I don't know. It can't, but you have to go out and get it or go to some effort or you're in a hotel room and you make a crummy coffee there and then you don't really enjoy it. It's just simpler not to have that addiction. I think...
What's very interesting is that caffeine pills are incredibly economically efficient. It's like one cent for the equivalent of a cup of coffee if you buy them in bulk and yet absolutely zero adoption. People love a cup of coffee, even if it's a couple bucks. It's like a simple pleasure. It's a joy. People just can't stop paying for them.
Do you do anything in the morning? Any type of ritual that you feel like you have to do to be yourself?
I get high from mineral water, but if I don't have it, I'm fine. I'm not craving it.
I'm addicted. Okay. Okay. Well, let's talk about artificial intelligence. How are you thinking about the latest debate around Kimi K3, open-source AI? Where should we begin the discussion? And then I'm sure we can narrow down a bunch of different ways.
Open source is coming and is here whether we like it or not. Of course, a lot of it will be from China. We should allow American competitors to proceed on an even footing. But basically, the attempt to outlaw it or ban it or use sanctions against it is going to fail miserably. Like what about multinationals? What about companies whose affiliates use open source? What about all the major US companies that use open-source Chinese products right now? I hope, I plead that the Trump administration gives up on this crusade.
Okay. So, I want to push back on it cannot be banned because I, at first blush, I agree with you. Like it is software. It is something that can just be downloaded. You can download the weights and just have it. Anything that's replicable infinitely should just proliferate across the world very quickly. But the fact that you need an NVL72, you need a whole server rack, you need equipment and infrastructure that acts as a point of leverage. I'm not saying that we should ban it. I'm just saying if you put me in charge of banning it, I think I would have an easier time banning open-source AI than banning music distribution, for example.
What will you do when it's on device? How long is that from now? Right. True. Why go down the path at all?
Yeah. Yeah. Yeah. Yeah. Yeah. Yeah. That's a good point.
Yeah. What lessons are there from the automotive industry for this moment? Just given, you know, the use of tariffs and how those have worked and haven't worked and we were talking, you know, earlier about how Europe's approached the problem. You see BYDs all over Europe now, but at the same time many of the manufacturers have been hurt most by just losing the Chinese market, right? Or not losing it, but having declining sales there because of local alternatives.
I think if Europe starts a major trade war with China, Europe will lose. This is a sad reflection on Europe and it's a sad reflection on Europe that they can't just say, 'Oh, we'll make fewer cars. We'll do five or six other great things that we're readily positioned to shift into.' They're not really there. But simply putting up big tariffs isn't going to help them get there. It actually will postpone it. And those tariffs over time end up applying to the inputs for your other sectors and it makes them less competitive. So Europe's really in a pickle. There's no good way out, but I don't favor the tariffs either.
How do you think about the open-source models in terms of like the broader AI stack? Like, is it important that America have some position in energy, data centers, chip manufacturing, networking? There's this whole stack and I've been surprised that we've been building as many data centers as we have in the United States given all the pushback. That feels like something that could have gone to at least other Western nations, energy-rich countries, but that's been fairly centralized. But there's clearly, it feels like there's a potential slippery slope for, oh, well, if China's handling the open-source models, then maybe they should have the data centers. Maybe they should also do the power generation over there. It's dirty. We don't like having data centers here. Put them over there. And then at some point it feels like you could lose leverage. But I don't know how risky that actually is.
I live right next to all the data centers in Northern Virginia. It's great. They pay for half of the taxes of their county. The energy, the imby, all that. I'm completely for it. The chips. Bring it on.
But at the end of the day, people are afraid that the American models can be yanked from them.
Sure.
That's partly our fault. Yeah. And if there's open source as something in the background, people are more willing to lock into our systems to a partial extent. So the two are complements, in fact. And again, there's all these major American companies. I don't typically advertise it, but they use a lot of Chinese open source right now. Yeah.
You just going to pull that out of their guts?
Yeah. Yeah. I mean, both enterprises actually deploying these models for cost savings and also, I mean, we saw with Thinking Machines, they used the Chinese open-source model to bootstrap a step in the process of building their latest product. And so it's a whole supply chain. Where do you think the data center pushback comes from? If you are there and have a positive experience, it's polling very poorly. Is it all fear around job displacement? Is it anything particularly on the ground? Like, where did this come from?
I think right now people are just negative about many, many things. They're negative about their country. They're negative about social institutions except the military. They're very low levels of trust.
They were told during COVID times, during the great financial crisis, everything's okay when it was not.
So you can't completely blame them. But look, as you know, this country has 50 states, and if Mark Zuckerberg needs to pick up the phone and call the governor of Louisiana, thank goodness that can work here.
And I think we're going to do it.
What is it about this? We're in this like counterintuitive vibe session moment. Kyler Scandlin sort of written about it, coined the phrase, this idea that when you poll people on how are your neighbors doing economically or financially, people say they're doing fine, but I'm doing terribly. Or the economy will grow, the job market will be stable, but people will report the consumer confidence is very low. I'm wondering if you've reflected on that more generally and then I do have a follow-up question, but what is the state of the vibe session? Is it just continual? Is it permanent? Is it getting worse?
I call it negative emotional contagion. And you see it in many other eras of world history. It feeds on itself. I'm not sure we know what starts it. Maybe it is some negative events, but once it gets going, it's hard to reverse. Sometimes you need actually a war or some kind of catastrophe. I hope we do not. But if you look at wealth accumulation, the job market, stock prices, real wages, many other indices, they're doing like okay to fine. They're just not doing terribly.
Yeah. No, no, I completely agree. And so I had this idea. I don't know if you'll agree, but maybe the fact that we're in this vibe session and the vibes are so bad and there's so much negative emotion around the economy, that if we actually were to go through a brief recession or a market correction, we might be a more resilient country, sort of ironically, because of that. Because everyone would look around and be like, 'Well, I already thought it was bad. It got a little worse, but I'm ready to reinvest. Maybe this is what I've been waiting for. Finally, I'm justified in thinking things are bad. And so I can get out of this and do whatever is necessary because there's no longer a cognitive dissonance between what the market is doing and how I feel.'
One hopes, yes, it could be a bullish signal. But then again, if something terrible does happen, everyone says, 'I told you so.' And maybe we get stuck in the mire.
About a year ago, you had a blog post, if I remember correctly, one of many, of course, that was something to the effect of like AI is real and it's amazing, but the economy is still going to just sort of generally continue growing as it's grown, maybe a little bit faster, hopefully. Are you seeing anything that could update your worldview on that? Because so far I feel like you've been pretty vindicated. You deserve a victory lap.
Well, I'm seeing a lot of confirmation. There's so many people that will say to me or say in chat groups or on Twitter, 'My goodness, if we had Fable 5, I would have thought the world would be totally different.' And it's not.
And that's my view. But look, you go back to 1995 through 1998, we got half a percentage point of extra growth out of those advances. They to me seem less impressive than AI. So, I think over time with some slowness, we'll get that and maybe a bit more, which is great.
Interesting. Have you seen The Odyssey?
Can you give us your review?
Well, look, compared to other movies, it's fantastic and it's dreamy and it's impressive, but mostly I was disappointed. I thought the dialogue was lame. The casting, I don't mind the cross-ethnic casting at all. I just didn't think they were the right people. Matt Damon is not Odysseus. He's just not impressive enough physically. The woman who played Calypso just seemed like a suburban housewife. I love Zendaya, but she didn't have any real role. So, they screwed up many things. It didn't quite cohere. I'm very glad I saw it. Everyone should see it. It's worth the three hours, but relative to expectations, I suppose a disappointment.
Do you think classicists are satisfied with it? Do you think it's good for the field?
They're delighted to have any attention at all and the different translations of the Odyssey are now on the bestseller lists. So, what's their alternative, right? They shouldn't complain too loudly.
Do you think technology is making soccer worse?
I don't watch soccer, but I think technology in some ways has made basketball worse. And I wouldn't be surprised if the same were true of soccer, that it homogenizes things and everyone chases after the same targets, the same players, teams have less individual style. But to soccer, I cannot speak.
What about AI maniacs? Take me through what defines an AI maniac, why it's important, and I have some pushback and follow-up questions.
I spoke to two teenagers this morning. One was 13, one is 17. Abu Dhabi, Finland. They're AI maniacs. They spend their spare time mastering AI. They know way more than their so-called teachers. They're starting companies, are building models. They realize this is the future. You're going to have a large number of companies with quite a very small number of employees but pretty high revenue. And I think AI maniacs will compete in medicine, law, business, consulting, banking, wherever it is we let them compete. So this to me is the future. Like if you're asking, will the UK recover? I would rephrase it. Is the UK going to encourage its AI maniacs?
How can they encourage these AI maniacs? It feels like AI naturally lowers the barrier to entrepreneurship. And so, there's not much encouragement that's necessary. I mean, as long as the models are available and affordable for young people, they will do things with them.
Young people should be allowed to rise to positions of authority, they should be respected. They shouldn't be looked down upon. They should have the chance to buy a computer, get grants. It's not automatic. So, you look at the United States, we do great on all these metrics. Estonia is a country that does pretty well on these metrics. A lot of parts of the world are too small-c conservative and we'll see where different countries are going to fall on this spectrum, but they're not all going to do it automatically, just like not all countries have made great rock and roll.
Sure, though. The technology is pretty straightforward.
Yeah. Yeah. Estonia is famous for major government investing in high-speed bandwidth. Is that correct? Is that the rough story of Estonia?
And but I'm interested in what that would look like for AI in America or the UK. Like it feels like many of the models or application layer companies, they have sort of a free tier that gets someone potentially from zero to close to AI maniac and then at some point they can go and find a customer to help pay for their compute bills. But is there more the government should be doing?
But as you know with venture capital, there's a valley of death. I don't know that the government should be making grants to 16-year-olds, but someone should. I'm doing it. Other people should as well. So this is really the province of youth. There's nowhere you can go to learn this stuff other than say some of the major AI centers and companies. So you've got to teach yourself. And is there the support in your environment for that? We will see.
Give us an update on how you're thinking about grants. Are you going younger, broader, smaller, different partners? I mean, there's been some folks who have been able to do compute grants and then they sort of cap, they sort of, the dollars go further because of the implicit margin.
With AI maniacs, I'm going younger and weirder. And I've also noticed Europe is doing pretty well on the AI maniacs metric. We'll see how it develops as the costs rise.
But I've been pleasantly surprised and I'm now more optimistic about Europe than I was six months ago.
Mhm. How does that, like where did that come from? That feels like a come-from-behind story for Europe. I feel like for the longest time the brightest technologists in Europe have gone to America to start their companies. The Collisons' famous example. There's many others with the exception of like the Swedes and Spotify, I suppose, but in general it feels like many people have come. Are they staying more frequently? Like what's happening?
Well, these kids may yet go to America. They're too young to be going now. Maybe it's that, you know, chatbots didn't interest the Europeans that much. They seem to like compete with culture.
Agents are a chance to do something. Now American kids have had opportunities to do things for a long time.