John Coogan and Jordi Hays host this episode of TBPN, which covers three main topics: a claimed AI-assisted proof of the Riemann hypothesis and its market impact, a deep dive into the stock market's negative reaction to AI news, and a review of a 500-year-old mansion. The hosts debate whether Wall Street is ignoring the significance of AI breakthroughs and explore the strange coincidences between Anthropic’s IPO and the launch of Grand Theft Auto VI.
> “It appears that there is a solution to the Riemann hypothesis, a proof, which is extremely consequential in the math community. Wall Street doesn't really care, apparently.” > — John Coogan
> “Silicon Valley says AI is fundamentally different. It is different this time. Wall Street team seems to say, you're a SaaS business with high capex. Not sure I like it that much.” > — John Coogan
> “It doesn't make economic sense. We have overcapitalized on the house. We have to sell everything. But when you are living somewhere, you want it to be done right.” > — Jordi Hays, quoting the mansion’s owner
Diet TBPN delivers the best of today’s TBPN episode in 30 minutes. TBPN is a live tech talk show hosted by John Coogan and Jordi Hays, streaming weekdays 11–2 PT on X and YouTube, with each episode posted to podcast platforms right after.
Described by The New York Times as “Silicon Valley’s newest obsession,” the show has recently featured Mark Zuckerberg, Sam Altman, Mark Cuban, and Satya Nadella.
TBPN is made possible by:
Ramp - https://ramp.com
Public - https://public.com
Cisco - https://www.cisco.com
Console - https://www.console.com
CrowdStrike - https://www.crowdstrike.com
Figma - https://www.figma.com
Lambda - https://lambda.ai
MongoDB - https://www.mongodb.com
NYSE - https://www.nyse.com
Railway - https://railway.com
Shopify - https://www.shopify.com/
Follow TBPN:
https://TBPN.com
https://x.com/tbpn
https://open.spotify.com/show/2L6WMqY3GUPCGBD0dX6p00?si=674252d53acf4231
https://podcasts.apple.com/us/podcast/technology-brothers/id1772360235
https://www.youtube.com/@TBPNLive
John Coogan
Let me also tell you how many days there are until Christmas. Because we're so close now. Time to do your shopping. I hope you didn't get 828 hours. I hope you didn't get anyone any stock for Christmas because it's going to be worth a lot less by the time they unwrap that under the tree. But if you bought the puts, they could be looking really good. Puts under the tree. Put some puts under the tree if you're bearish this holiday season. It's over. It's over. The market is nuking. Pure red. Everything is nuking. Anthropic is nuking. We got to add. So we got to add Anthropic to the ticker.
Jordi Hays
Yeah. Ant at the bottom there. Throw it in there.
John Coogan
But it's a whole. It's still a big company. And they got a lot of cash. Doing well. But people are frustrated. Market cap's $1.14 trillion. That's, of course, the active, not the fully diluted. Stock's at $87, which is still above the $83 that the IPO at. So there's a little bit of a bump since an hour ago, but down 10% today. And of course, everyone who loves to read into the accelerationism that comes from math accomplishments is confused because it appears that there is a solution to the Riemann hypothesis, a proof, which is extremely consequential in the math community. Wall Street doesn't really care, apparently. Chad is saying sell the dip. Sell the dip. Levant over at Anthropic posted that he believes he found a proof with Claude's help, but Scientific American reports it as unverified and still needing expert review. That could take years. It's 250,000 pages. Is that right?
Jordi Hays
Yeah, I think that was the raw part. I mean, you can definitely trim these things down. That's crazy.
John Coogan
That's crazy. I mean, we're having Eric Winn Vintage on the show today. His book is substantial, only 350 pages. Imagine 1,000 of these books, maybe one day. With AI, it might be possible. And so the Clay Institute still lists Riemann Hypothesis as an unsolved Millennium Prize problem. So all we're hearing is the claim. People are going back and forth on it. And Terrence Tao was on Tyler Cowen talking about the importance of the Riemann hypothesis. Also, he went on a show called Numberphile, and he says he believes that, he said he believed with the help of AI we could solve the Riemann hypothesis before 2050, and it took 35 days, if this is true. But he gives some good context here on why Riemann is important to the math community. It's sort of the thing that everything else is built on. So let's play this clip of Tyler Cowen talking to Terrence Tao.
Tyler Cowen
Did not accept what the math was telling them. And the Riemann hypothesis, what do you feel?
John Coogan
Oh, yeah, that is a really big prize in mathematics. And actually recently we're getting some progress on that. That's another story. Pause. This is funny because Tyler's original question was, you don't need to tell me about the individual math problems. I'm going to hit you with a sequence, a lightning round of math problems. I want you to tell me how you feel about them. What emotion does it conjure? Now let's listen to Terence Tau's answer.
Terrence Tao
It concerns one of the most basic objects in mathematics, the prime numbers, which we now use prime numbers all over the place, like in cryptography. There's a lot of cryptographic protocols that we use to encrypt our finances and whatever, rely on properties of the prime numbers. And we look at the prime numbers, and they behave very randomly. Sometimes they're close together, sometimes they're far apart. And the Reitman hypothesis is a very precise way of quantifying this random-looking nature of the primes. And it has many, many consequences. Is that any part of mathematics that touches the primes, if you know the Riemann hypothesis, normally your results get stronger. So there's a large chunk of results in number theory which are conditional. If you have the Riemann hypothesis, you get this good result. Otherwise, we can't say anything.
John Coogan
So not exactly an emotion, not exactly a feeling. Sort of dodged the question there. But very, very informative answer, nonetheless, about why this is important. Basically, post-Riemann hypothesis, not proof. A whole bunch of mathematicians did a lot of work that said, okay, we're going to assume that the Riemann hypothesis is true and that the prime numbers do exhibit this behavior. And then we're going to go do other math that assumes that this is true. And we're going to hope that in, you know, 50 years, 100 years, it's not proven false. And today, potentially, this could be serious history.
It is proven true. And therefore, everything else holds and everything else becomes true. So if you've been doing work in math that assumes Riemann was going to be true, it's a huge day for you because you didn't get disproven. You can build on top of that successfully. So very, very good. Anyway, AI capability, it's obviously getting better at a steady pace. You feel this with all the new models, all the new capabilities, all the different benchmarks, even the vibe check benchmarks, the 3D modeling, the video game building, the AI video editing, everything that you see. You can see the progress month to month, but the market doesn't care.
The market does not like it. Anthropics down 10%. The NASDAQ and S&P are way down as well. Nim Chimski says the Riemann hypothesis is worth approximately negative $121 billion, because that is, of course, the drop, the pre-market drop on Anthropic. But these things are obviously unrelated. People are not trading public market stocks based on math progress. But there is a question of when or will the market digest the RSI narrative? Silicon Valley has come around on that idea. People are reading into the Meter Report, the Anthropic Model Cards, around AI development speed up. Is there a compounding advantage here?
There's this back and forth between. Silicon Valley says AI is fundamentally different. It is different this time. Wall Street team seems to say, you're a SaaS business with high capex. Not sure I like it that much. But obviously, the folks in Silicon Valley say, no, this is wildly different. RSI will create, if anything, it'll be too permanent of a monopoly. We could risk a singleton. We could risk a takeoff. These are all extremely bullish signals if you believe them and you believe that you can capture value over time in that regime, which is a whole other question. But at the moment, it seems like the market is reacting just to overall weakness in the stock market broadly.
NVIDIA chaos, geopolitical chaos, cyber attacks chaos, all things that would stop customers from actually forking over hard-earned dollars to move the needle on revenue, on profit, on the things that Wall Street likes. And they not really focused on the math results just yet.
Jordi Hays
But maybe soon.
John Coogan
Yeah.
Jordi Hays
And the last couple of weeks have been looking at your phone seeing red and then seeing that there was another hack.
John Coogan
Yep.
Jordi Hays
And then seeing some more red another hack. So Derek Thompson summed it up nicely. He says NASDAQ up only 5 year over year. It's all over. And we will see. We'll see where the year ends for the NASDAQ. Year to date, The NASDAQ has underperformed T-bills. If at the start of this year someone said you should just be holding bonds, they would have been correct relative to tech stocks, which is not what anyone predicted. At the beginning of this year, remember, we were talking about the SaaSpocalypse. Would there be a massive reallocation of value and investment from SaaS companies to AI companies to bottleneck companies?
Everything that's happened with the booming bottleneck stocks, the energy infrastructure, the data centers, the chips, every piece of the stack, they're all in the NASDAQ. You add all that together, you still underperform the U.S. government and the T-bill. Seriously. Michael Burry says, CoreWeave, Nebius, NVIDIA, Semis, I tried to warn you all. I didn't say short, but I provided the information one would need to sell. We should start saying that because we talk about every company. Every time a company goes down, we should say. I didn't say short, but I provided information. But it's funny after this historic run-up, then just being like, you know, there's people sitting around that did take profits.
So to be clear, Michael Burry, on January of 2023, posted just sell with a period. 2023. And there were a lot of ways to make money between January of 2023 and November of 2026. It has been a lot of activity in the market, a lot of gains. And so people are saying, I mean, the S&P is close to doubled since 2023.
John Coogan
Wow. Since since that post, it's hard to hard to see. Exactly. Anyway, whether you're long, he might have posted the next day by just didn't get the engagement. It's over, Agent. Sell everything. SpaceX. There's news about SpaceX. SpaceX leaders want to – lenders, not the leaders, lenders, want it to stop – want to stop SpaceX from launching $40 billion worth of AI chips into space, according to the Financial Times. The chips back a roughly $40 billion financing deal led by Apollo. An Apollo-managed entity buys NVIDIA GPUs and leases them to SpaceX, which pays them back over five years. If SpaceX defaults, the lenders can take back the chips and sell them. That's the problem with space. Lenders worry that if SpaceX sends the chips into orbit for its planned space data centers, they couldn't reclaim them in a default. You just need a space tow truck. This is a new business. You need a repo man for space. Kind of pathetic from Apollo.
Jordi Hays
Yeah, you should just launch your own rocket to go drag it down.
John Coogan
Yeah, just say you're not getting into the launch business yourself.
Jordi Hays
Yeah, Apollo should get into the launch business with a tow truck for space. Don't yeet all of our collateral into space, please.
John Coogan
No, that actually does change the economics because, yeah, there's been so much reallocation of compute, I mean, down to whole projects changing hands from one hyperscaler to another, leases and permit approval. We have Alex Epstein coming on the show from the Center for Industrial Progress at 1130, and he's going to help us understand some of what's going on in energy around permitting. And it's very interesting because the land can change hands three times before something, and then the actual shell can change hands. The power rights and the water rights can change hands a bunch. Then the company that's building it can change hands because someone didn't have the right staff. And then the actual chips that go in there can be reallocated. Semi-Analysis had an interesting post about how some older data centers, the floors just aren't strong enough for new NVIDIA racks. So that's sort of because the new Blackwell racks weigh, I think, three times or five times as much as they used to. And so you have to go and retrofit the building if you actually want to bring in the latest and greatest. So there's very...
Jordi Hays
Here's a question. So Trump has been talking this morning saying that he's not going to allow data centers outside of the U.S.
John Coogan
Yeah.
Jordi Hays
So it's kind of interesting. You have, one, some U.S. politicians saying no data centers in the U.S., Trump saying no data centers outside of the U.S. We'll see where they land. But how do data centers in space fit into that? Kind of a perfect loophole. I feel like it satisfies both.
John Coogan
But he's saying he's only allowing you to build data centers. Well, you're building the data center technically in the U.S., and then you're launching it into space. Outer space. I think that counts.
Jordi Hays
Okay. So Elon has been a huge fan of Backwards Day for a long time. He has said Anthropic is destined to become misanthropic. Open AI will become closed AI. Mid-journey isn't mid. He's got zingers. But he says it's impossible to come up with an opposite-day zinger, an opposite-day epic vulgar roast about SpaceX, because what's the opposite of SpaceX? It doesn't make any sense. What's the opposite of X? Why? That doesn't make any sense. But he's wrong. He forgot about ground X, which is what's happening to SpaceX because SpaceX is making more money on the ground than in space now because of the data centers.
So it's becoming ground X. And they also added AI. It was SpaceX AI. And then they changed that to SpaceX SI. And we all know that there is an opposite of super intelligence, midwit intelligence. So it could be ground X MI. If things don't go well. But I think he can launch these into space. But we'll see. It is a very funny opposite day. The zingers. I love an epic vulgarist. He's a good guy. He's the best. So SpaceX closed the deal in October with the leases set to begin in the first half of 2027. Musk had said the debt markets have almost entirely shut off access to money for AI companies without other big revenue streams.
John Coogan
So, yeah, you have to be able to underwrite your loans, your debt with your cash flow. For a long time, the hyperscalers didn't really have anything to do with their cash flow. They would just build up on the balance sheet. They would buy back their stock, maybe pay dividends. Now they're able to use that cash flow to finance big debt offerings. But it only really applies to companies with large cash flows. Meta, Google, Amazon, Apple. Although Apple's not really tapping the debt markets just yet. But SpaceX is in that category because they have Starlink. That generating billions of free cash flow.
It not in the hundreds of billions a year of free cash flow like some of the hyperscalers but it certainly allows him to pull this $40 billion financing together as long as he keeps the data center. What kind of run rate is Starlink on? I thought it was like $10 billion. I don't know. You can go pull it up. I remember when SpaceX went public, people were surprised that the business was like a little smaller at the time because launch is such a small category. And then Starlink was still ramping and doing very well, but it was small relative to $2 trillion. But, of course, you're investing against, like, Elon and the long-term vision.
You're not just buying, like, a telecom, right? You can pull that up. Working on it. So, what did you get?
Jordi Hays
Starlink's revenue run rate is about $21 billion a year.
John Coogan
That feels high or low?
Jordi Hays
That's annualized off of Q3, 2026.
John Coogan
Yeah, that feels right. I mean, they're in a bunch of planes now. They have big deals with the airlines. Tons of people have Starlink. I have a Starlink. It's great. And they're starting to direct to sell, but there's some problems with that. It won't fully be able to replace sell because it doesn't work indoors everywhere. Also, there's just this interesting dynamic where the telecom companies have spent decades finding little patches of gaps in the coverage. So you used to go to like a basketball game or a football game and you would lose connectivity because there were just so many cell phones in one area trying to break through, trying to get connectivity. The networks have figured out how to load balance that and actually installed extra infrastructure and extra backhaul to make sure that you can maintain connection while you're in a crowded group of people. If it's expected. Sometimes if you're in a place that no one expects there to be 100,000 people all of a sudden, you can get sort of jammed up still. But that's a lot harder with SpaceX because with the Starlink network, they still have to go and do that work to figure out how can it be reliable. What are you laughing at?
Jordi Hays
I'm laughing at Strauss's response to my question yesterday, which 82,000 people have appreciated.
John Coogan
Oh yeah yesterday i asked if gta 7 you know we were having a conversation about you know the the possibilities of generative ai and in future versions of gta asked if gta 7 could just be set across the entire world and he said we just went live with gta 6 on november 19th oh it's still november 19th and you're asking a question about gta 7 that we we we clearly thought that was much more funny than he did but we had a great conversation with him yesterday he's also it's also funny because we we will make we will use a giga chad filter on various guests of the show but this is i think just a real picture which is a work of art.
Absolute Chad. Well, Nabeel and Good Alexander are going back and forth on Riemann. Claude solved the Riemann hypothesis and Ant is down 10%. Seems like the market is not RSI-pilled and believes in AI as normal technology. Good Alexander asks a question. How do you make money off of solving the Riemann hypothesis? Who pays? Well, the Millennium Prize pays a million dollars. So you should add that to the cash flows. And it should hit in a couple months. You don't have to discount it back that much. So you've added maybe $800,000. You might even be able to get a financing against that million dollars that's coming in when the money hits.
So that should add to the book value of the company. If you're investing in Anthropoc on book value, this is good news for you. But who pays in general? No one's really buying hypotheses these days. Nabil says, yeah, the steel man of this case would be something like this. This is way faster than anyone expected, even conditional on the fast math progress earlier this year. Something similar will happen for AI research. Having an automated AI researcher ahead of competitors is going to enable various ways to make money. So still like some leaps there, but that is the RSI trade. And I do think people will wake up to this.
And you could even say that they are because it's a high revenue multiple, high earnings multiple for a company. Obviously, it's a young company. It's high growth. But it's a lot of fun. We were digging into the question of, is there more than meets the eye to the fact that Anthropic went public on the day GTA VI launched? Like, are they trying to tell us something with that? Hidden messages.
Jordi Hays
Yeah, there's no coincidences in this world. And so I dug into it a little bit more, and I think there's something going on here. So in 2001, Claude Shannon, the father of information theory, died. That's the namesake for Claude, the model for Anthropic. And that same year, Rockstar introduced Grand Theft Auto 3. And Tyler, can you look up who the main character, who the protagonist is in Grand Theft Auto 3? What's his name? Claude. It's just Claude. Interesting. Okay. So 25 years later, February 24, 2026, exactly the 25th anniversary of Shannon's death, Anthropic released version three of its responsible scaling policy.
The company's official framework for preventing catastrophic AI risks. Then on May 26th, and everyone knows what May 26th was, that was the original scheduled release date for GTA 6. So we're already seeing more connections to GTA 6. Tyler, so May 26, GTA 6 was supposed to launch, May 26 of 2026. Tyler, I want you to look up version 3.3 of Anthropics' responsible scaling policy. Pull that PDF up. The responsible scaling policy, version 3.3 specifically. And this is going to shock you. I want to see the date that that was released. May 26. May 26. That's crazy. Rockstar subsequently delayed GTA 6 how many days?
How many days between May 26th and November 19th? I want you to look that up. 177. 177 days. And do you know the value of Anthropic when it went public, the precise value? $1.77 trillion. $1.77 trillion. The connections don't stop there, though. Both companies were founded by siblings Sam and Dan Hauser at Rockstar Dario and Daniela Amadei at Anthropic. And get this, Anthropic priced their IPO at a share. Guess what year Dario Amadei was born? 1983. 1983. Yeah. I'm going to blow your mind. That's a similarity between two Anthropic things, not even Anthropic and GTA. Tyler, you're telling me we're seeing patterns that don't exist?
Wow. Okay. Okay. We're not crazy.
John Coogan
You're crazy. You want a real connection between GTA and Anthropic? I'll give it to you. You know Reed Hastings? He's on the board of Anthropic. You know what company he founded? Netflix. You know where you can stream GTA 3? Netflix. That's right. Yeah. Yeah. Netflix is a distributor for Grand Theft Auto. So the guy who's on the board of Anthropic just happens to be distributing the game through Netflix at another company. Interesting. Chad is saying Tyler's trying to cover this up. He doesn't want the truth to get out. Okay, I want you to rearrange the letters, Anthropic, because it's a perfect letter-for-letter anagram of hip, cop, ran. Coincidence? I think not. We're going to pull up that image. Let me send this in. There's more. It goes deeper. Where does GTA V take place? San Andreas? No. Oh, there's a city. Los Santos. And the anthropic ticker is hiding right in the middle. S-A-N-T-O-S. Los Santos. What a coincidence. I think that's it. I tried to find as many. There's definitely more.
Jordi Hays
There's definitely more. Jack Clark is also an anagram for carjack if you leave out the L in the day. Coincidence? I think not. I think not. Symbolism will be their downfall. Let's go over to the mansion section. We've got to do some tier lists. Jordy, I want your review of this 500-year-old mansion. They wanted a bigger house. They ended up with a 500-year-old mansion. They bought a 12,100-square-foot tutor. They had to rehome protected bats. There you go. If you move into a haunted mansion, right? If there's bats in there already? Not necessarily. No? You think the previous owner was just keeping pet bats?
These are friendly domesticated bats? Well, I mean, people are afraid of bats for good reason. Yeah, rabies. Get disease from them. But, you know, some people actually train them and sort of get them to the point where they're on their side, right? Like witches. I thought you were talking about Batman. Oh. Yeah, that too. Can Batman control bats? He's not like Spider-Man where you can talk to bats, right? Can't Spider-Man? Spider-Man can't do that. There's someone who can. Aquaman. Aquaman can talk to the fish. Yes. Spider-Man was bitten by a spider, but Batman was not bitten by a bat. He just likes the vibes of bats.
Yeah, yeah, yeah. He's just like aura farming. He's just aura farming. Yeah, 100%. Joanne Hughes and Philip Stapp wanted a bigger house, large enough for their blended family of seven children, with a spacious backyard for their youngest, Billy. They ended up with an almost 500-year-old Tudor mansion, four times the size of their previous house. On nearly 13 acres in England's Shropshire Hills, the bat-filled property had only one shower when they brought it. Shropshire? 13 acre, 12,000 square feet, one shower. Brutal. I went to Shropshire and no one knew you. Everyone knew you. The couple embarked on a seven-finger makeover that is still in process several years later.
He's a 64-year-old financial advisor. Seven finger makeover? You said finger. I misspelled it. Do they mean like an AI makeover? Yeah. Probably started there. He's a 64-year-old financial advisor, and he admits they may have gone overboard after falling in love with the property, which now has eight bedrooms. It doesn't make economic sense. We have overcapitalized on the house. We have to sell everything. But when you are living somewhere, you want it to be done right. Before the move, they were living in a 3,000-square-foot, five-bedroom house in the city of Wolverhampton. Some 130 miles northwest of London, accommodating their children, now aged between 10 and 39 at the time was a squeeze.
Billy was a toddler at the time, so the backyard was also a problem. And look at what they did. Wait, their children have a 29-year age? Well, it's a blended family. So I think they both had kids and maybe had more kids together. So they've had a lot of kids across the two of them. Step always admired Shropshire's Hills, a 400 plus square mile swath of protected countryside around 400 around 440 miles west of wolverhampton he was the one who spotted i swear they're just making a name they're just messing with it listing for it's called place hall p-l-a-i-s-h uh the 12 000 square foot home was asking 2.
69 million dollars i feel like turn this into a hacker house get 50 people in there split it up you're living pretty pretty nicely yeah you're gonna make it all back you make it all back the hall was built between 1540 when king henry the eighth was in the latter days of his reign and 1580 by the wealthy they don't even know they have a 40-year range for when this house was built they're just like when was your house built it was it was built in the 16th century that's all they got by a wealthy judge and politician sir william layton according to a conservation group historic england the h-shaped house was built with red brick and unusual and very modern this is this is one of the best homes we've ever had in the mansion section you get this s tier uh this is yeah i'm giving this s tier is it are they going on the nice list so much character yeah definitely on the nice list but like look at like the fireplace kitchen everything is done they did a really good so nicely yeah it feels like it would constantly be cold in there is the only is the only thing is my only knock on it but they've added and imagine being imagine being the 10 year old just like you know trying to figure out who you are pacing around yeah or a farming they're just yeah born in the aura farm for sure yeah 130 acres too that's a lot of room anyway and that's our show thank you for tuning into tbpn today sign up for our newsletter good at tbpn.
com. Good. Leave us five stars. Good. On Apple Podcasts. Good. Buy Eric Wynn's new book. Buy Eric Wynn's new book. Make it a New York Times bestseller. Book a reservation in Mamafuku. Good. Have a great weekend, everyone. Have the best weekend ever. The market can't stop you. Bye. Bye.