Starship Launch, World's Fair Retrospective, Matt Grimm Does the Murph, Big China Moves

I see a large IPO on the horizon. You're surrounded by jungle. Hold your position.

This misinformation clearing order inbound. That's just wrong. We are surrounded by jungle. Hold your position. Come on, get up. Trust the experts here. We are experts. Founder knowledge.

I see multiple journalists on the horizon. Stand by. UAV online. Blaze win. Double blaze. Triple blaze. Double kill. Finally, it's a win. Team Deathmatch. We are experts. Triple blaze. That's just wrong. Right. Hardy clearing order inbound.

We are surrounded by journalists. Hold your position. Strike one. Cortisol sponge. Strike two. Cortisol rising. Activate golden retriever mode. Market clearing order inbound. UAV online.

You're watching TBPN and today is Friday, May 22nd, 2026. We are live from the TBPN Ultra Dome, the temple of technology and not only the fortress of finance, the capital of capital.

Great show for you today, folks. It's Friday, Friday, May 22nd. We had our Wall Street Journal stolen from us. We only have the Financial Times today, but we're making do because we have a digital subscription as well. And we're going to take you through America's IPO boom. We're going to talk to Matt Grimm from Andera. We're going to talk to Dan Shipper from Every. We're going to talk to James from Appeal Sciences. We're also going to read through Brandon Gell's news roundup in the TBPN newsletter, TBPN.com. You can go sign up. So, three key stories that broke since we last spoke a mere 21 hours ago, since we podcast for three hours and then we come back 21 hours later.

We do it again. We do it again. Yesterday, Politico reported that David Sacks made an 11th-hour appeal to President Trump to spike an executive order that would have created a voluntary program, voluntary program, for frontier AI companies to submit their models to the government for review 90 days before new model releases.

I wonder what the benefit of that is. Obviously, it's good to have red teaming going on, good to have benchmarking, good to have other eyes on the project. There's some sort of, you know, liability that happens. What's the round thing on the table? No one in the TBPN audience has ever seen this. Can you describe what this is? This round-shaped argument.

Not exactly sure what we're looking at, but I saw them talking about this on ESPN, and people are always saying, 'You guys are like SportsCenter for business.' You should pick one of these up. This object, so I had to check it out. Yes, this artifact. Brown, brown oval object. Yes. Cool. Well, it looks fun. Anyways, I was having a little bit of fun earlier throwing this drill in the gong. Drilling the gong. The production team didn't like it because there's this, you know, $70,000 massive flat-screen TV next to the gong as well as multiple cameras and lights over there. But no accidents. Wow. Mandate of heaven over there. Right here. Oh, Tom's got the tennis ball. Get out of here with that tennis ball.

Anyway, so for his part, Trump explained the last-minute cancellation by telling reporters, quote, 'I didn't like certain aspects of it. I think it gets in the way of — we're leading China. We're leading everybody, and I don't want to do anything that's going to get in the way of that.' The presumed text of the EO was leaked today. You can read it through this link. Tyler, have you checked in with LessWrong? Yeah. How are they feeling about this? Yeah, I mean, I assume they're not happy with this. This seems to be like anti-safety.

Do they want more review from independent nonprofits, for-profit companies that are run by people with good intentions, or the government, like democratically elected? I mean, I assume at some level it's got to be the government, right? Because if you're super hyped, these things are nuclear weapons. It needs to be the government. Yeah. But I think on the way there, people are definitely like, is there anyone who says like, 'No, no, no, actually, like nuclear weapons are super important, but they shouldn't be regulated by the government because I don't like the government. I would rather my favorite megacorp owns them, controls them.' I don't think there's — I mean, if you look at the popularity, if you look at the approval rates for different organizations, there are plenty of companies that are polling above the US government. Like people love Disney. Disney adults, they say give Disney a nuclear weapon. They already have the rights to build a nuclear power plant. You know this?

I did not. Disney World has like, you know, legacy rights because when they set up Epcot, which is an acronym which I should look up because what does it stand for? It stands for something really cool. This also bridges into one of our future segments. Stands for — what does Epcot stand for? Experimental Prototype Community of Tomorrow. The idea was to build an entire city. It was the practice of the — Yeah. And so in 1966, Walt Disney conceived it in 1966 as a utopian, fully functioning city. And I believe, this might just be viral clickbait, but I've heard it said many, many times that they have the permission to build a nuclear reactor to power the Experimental Prototype Community of Tomorrow, Epcot. Yeah. And so I don't know, maybe there's some folks out there that say, 'We can't trust the government. We can't trust the Democrats. We can't trust the Republicans with the nuclear weapons. Give them to Disney.' I don't know.

I trust Bob with the new — I would trust Bob, the new guy. We got to meet him. We got to get to know him. That's right, Bob. Bob, he can handle it. Yeah, I trust him. I trust his hand on the button. Anyway, what else is going on? Second story highlighted by Brandon Gell in the TBPN newsletter. The Gellinator in the TBPN newsletter. An anon account posted yesterday. He's sitting here with noise-canceling headphones. He doesn't know we're talking about him. He's going to find this out later when he watches these clips. Anyway, an anon account posted yesterday that Microsoft had canceled its internal cloud code license after token-based billing made the cost untenable. This catalyzed a ton of commentary and discourse. Box CEO Aaron Levie, for example, used this post to argue that token price optimization is set to become a prevailing trend inside companies relying on LLMs. Worth noting though that there's a proposed community note on the post right now that denies Microsoft made the move because of a cost issue. They have the money to spend billions on tokens if they want, but instead they are trying to shift the developers to use their own Copilot CLI, which I imagine they're going to build their own harness, their fork of cloud code. They want to make the engineers eat dog food.

Yes. But what's the dog food made of? It could be made of GPT-5.5 tokens. It could be made of Opus 4.7. I do think it — I guess they will be using — Yeah, I mean, I think the reason that Codex and Claude Code and Cursor and these other tools have gotten so good is because the engineers building them are using them all day long. And it's one thing to use those tools to build your coding tool, but being forced to constantly use the product that you're shipping is a tried-and-true way to build a great product.

Yeah. The third story from Brandon Gell in the TBPN newsletter: the American semiconductor manufacturer Micron announced that they've begun manufacturing 1-alpha DRAM, the most advanced memory ever produced in the United States, at its Manassas, Virginia headquarters or manufacturing plant. This move comes in the middle of the agent boom in AI, which requires longer context windows and thus memory than back-and-forth LLMs, which has led to an industry-wide memory shortage. Micron has been on a tear. Let's see what they're doing. Almost $850 billion market cap. Remarkable. It's up 2% today, 3% over the last five days, and 246% over the last six months. What an absolute run for Micron. It's up almost 1,000% over the past year. Really, really incredible performance from Micron. Just wow.

Anyway, America's going through a mini — This was such a funny article by the Wall Street Journal. America's IPO mini boom. How is this a mini boom? You got three of the boom. This is a giga boom. So, I don't know if we have to truth on this, but it also made it to the front page of the Financial Times. They said SpaceX, OpenAI, and Anthropic IPOs set to ignite Wall Street trading frenzy. NASDAQ loosens rules. Passive investors could dump rival stocks. Trillionaire prospect for Musk. The blockbuster listings of SpaceX, Anthropic, and OpenAI are set to prompt an unprecedented wave of buying and selling as new fast-entry rules thrust the tech stocks straight into Wall Street indices. The rules implemented this month by NASDAQ mean dollars of passive money will automatically flow to the three companies when they go public, driving their share prices higher and forcing investors to sell rival stocks. Elon Musk's SpaceX initial public offering next month is expected to be the largest on record, firing the starting gun on what US bankers will hope will be a blockbuster year, with OpenAI filing its IPO paperwork as soon as this week and Anthropic planning to float its shares too as well. The trio is set to raise tens of billions of dollars at a time of relentless investor appetite for companies linked to AI. The SpaceX listing will cement Musk's control of two of America's most valuable companies, SpaceX and Tesla. It is crazy. We're going to need a new term for the Mag 7 because Tesla has been in the Mag 7. It's been north of a trillion dollars for a long time, but SpaceX is going to go out at a higher valuation almost certainly. Where is Tesla trading at today in terms of market cap? It's at 1.34. Hard to imagine SpaceX trading below that. So, you're going to have his Mag 7 company will be lagging his non-Mag 7 company. We're going to need new terms. We're going to need a bigger boat. But why did the Wall Street Journal call this a mini boom? I think it's just because it's so concentrated on these three companies, but we will dig in and see where it goes. So, too bad SpaceX and others didn't go public sooner. I agree. But they are a tribute to the US capitalist system, says the Wall Street Journal. Today's AI investor euphoria recalls the heady dot-com days of the 1990s, and it's hard to tell how these bets will shake out. SpaceX, OpenAI, and Anthropic are rushing to go public to take advantage of the favorable market conditions. But amid the eulogies written about American capitalism, the latest mini IPO boom is a welcome tribute to the dynamism of US markets that no other country can match. SpaceX fired the starting gun. They're using the same language as the Financial Times over there. The starting gun has been fired. It's unanimous across the mainstream media. The rocket company founded in 2002 is a global leader in space exploration. Overnight success. It has also spread into broadband, mobile satellite internet, and data center development. Anthropic recently agreed to pay SpaceX $1.25 billion a month to use its data centers to train models. That shows how competitors can enjoy symbiotic relationships. SpaceX says the company as a whole reported a $4.9 billion net loss last year, some of which stemmed from its X.com social media platform. Can that be that much? I'd have to imagine that the losses were not — I mean, I don't know how much Twitter was losing before the take-private. I know that the revenues cratered a bit, but there were also big layoffs and an 80% reduction in cost. I think the word 'some' there is maybe doing some of the lifting. I imagine that the losses were sort of spread over all sorts of investments.

You're building a new Starship rocket which is set to launch today, 5:30 Central time. Tune in. The last one was — what a critical launch. Is it? No. Just because of the timing with the S-1 and the IPO. I don't know. All eyes on — I think the market is valuing the launch capabilities and SpaceX and Starship in particular is like one small piece of the puzzle. You know, it's important, but people are going to be paying way more attention to this. Yeah. Obviously, people will be excited about it, but I think even just the endless spamming Falcon 9 can get Starlink to what, a $20 billion run rate or something, a $10 billion run rate. You can clearly build a business off of the existing capability. Lots of excitement around Starship. There was a little bit of spice in the timeline from Everyday Astronaut because he was sort of rooting for a scrub and everyone was like, 'Oh, he's against SpaceX.' But he was like, 'No, I'm such a fan that I wanted to be there in person. I couldn't be there in person, so I hope that they scrub because I want to be there for this historic moment.' So a little bit of a misconception there. And the live stream of the launch is, you said 5:30? 5:30 Central time is like when it's scheduled, but they're already counting down because they stacked it yesterday. They were ready to go. Nicki Minaj was on site talking about Starships were made to fly, which I believe is one of her songs. SpaceX said in its filing that it foresees $28.5 trillion — yes, trillion — in market opportunities, including data centers in space. Is anything more bullish than the future just being enterprise software? Like you paint this massive picture. We're going to Mars. How are we going to make money? Enterprise software. $22 trillion of enterprise software. We're going to be automating workflows for the enterprise. I guess forever. Mr. Musk in January was awarded 1 billion performance-based restricted shares that will vest if his company, quote, 'establishes a permanent human colony on Mars with at least 1 million inhabitants.' Do robots count as inhabitants? Do agents count? Isn't that one of the potential — You got to get a million. This is why he's — Wasn't there a list of like five or six things and he just has to do like four? I don't know. I haven't dug into that particular one. I know for Tesla there were a whole bunch of different milestones. The humanoid robots was one of the stretch goals. But permanent human colony on Mars with at least 1 million inhabitants. That's exciting. That's a good goal. That feels like a whole lot of enterprise software and a million people living on Mars. That's a vision I can get behind. I'm into that. I like it.

Anyway, we have Matt Grimm from Andera, the co-founder and COO, joining us. We're calling in to the Murf. We talked to him last year. Let's see how he did this year. Matt Grimm, how are you doing? Welcome to the show. I can't hear them. You can't hear us? Oh, no. Okay. If anyone can hear us, just ask him to talk. Pass it along. Just have him rant. Just rant. Tell him to rant. I'll text him. Rant. Rant. We might need to come back to him, but he does look like he did it in the weighted vest. He's looking very sharp and it looks like it's cold and rainy. This looks brutal. I'm excited to come back to that in a minute, but in the meantime, we will go back to government pension funds and the SpaceX IPO. Government pension funds cave that the IPO will preserve Mr. Musk's control of the company by giving him more votes per share, but dual-class structures are not uncommon. Berkshire Hathaway and News Corp also have them and can encourage founders to take their company public since they won't lose control. If pension funds don't like SpaceX's structure, they don't have to buy shares. SpaceX could fetch a market valuation of upward of $1 trillion, a huge boon for its early private investors and employees with stock options. OpenAI and Anthropic also have heavily compensated their employees with stock options, partially as a way to preserve capital. Stock options allow employees to share in a company's growth and success. Thanks for the context, Wall Street Journal. I appreciate that. The shame is that retail investors have missed SpaceX's early stratospheric gains. What was the market cap of Tesla at IPO? What was the market cap of Tesla at 10 billion? Right. Was it Tesla's — Tesla — 1.7. Whoa. 1.7 billion. Yeah. That is — is that a market cap for — You got a thousand-x if you just were a day-one IPO believer. I mean, that's how you build a retail army. So SpaceX, I mean, a million people on Mars, we could be looking at 1.7 quad potentially. Anything's possible. I don't know. You get a thousand-x from here if we get to Mars. That's a big, big opportunity. That is so crazy that you could buy an Elon company in the public markets. Yeah. At less than a $2 billion market cap. And many people did not. I owned a little bit of Tesla at some point, but there were so many moments when it was so wildly disconnected from other car companies. It was never a Warren Buffett buy, you know? There were always good reasons not to buy it. Instead of going to high school, I just bought the Tesla IPO using the money I made from mowing lawns. Yeah, you should have just been a lawn mower who DCA'd into Tesla the whole time. Probably be pretty retired right now. I mean, there's a lot of people that did that. There's also a lot of people that put money down to buy cars and then they didn't buy the stock and they're like, 'Ah, the car depreciated and I should have bought the stock.'

Anyway, let's check in with the Starship launch because SpaceX postponed the launch of the newly redesigned Starship, but hardly a headline if it's just a one-day postponing because these things happen all the time. But it's completely redesigned. It's 400 feet tall and it stood on the launchpad at the company's Starbase complex outside Brownsville, Texas. But engineers had to troubleshoot problems that kept cropping up to the end of the countdown for the flight. According to a SpaceX live stream, SpaceX chief executive Elon Musk later said in a social media post that a hydraulic pin on part of the launch tower didn't function correctly. The company could try the flight Friday if the problem is fixed overnight. So, they worked all night and they fixed the hydraulic pin on the launch tower. Wow, talk about complex. The test mission was set to be the first flight for Starship V3. This is the third version of the redesigned SpaceX vehicle intended to deploy bigger satellites and one day help the company potentially settle Mars and mine asteroids. Starship is bigger and more powerful than any rocket ever built. The company has sketched out a future in which Starship flies thousands of times per year. Of course, the Mechazilla tower that grabs it, the chopsticks, that's very key to being able to stack these quickly. You can't rebuild all the infrastructure. The reusability is key to flying regularly. The rocket has yet to fly any customer satellites after three years of intense tests that we know of. The real conspiracy theorists — there's conspiracy theories that think space is fake. Nothing's happening. They're not actually going. The satellites don't exist. And then there's other conspiracy theories that think, oh yeah, when they crash, they're actually deploying secret satellites, and there's more stuff in space than we think. So, pick your tinfo hat theory. We'll let you be the judge of what's really going on. SpaceX's growth plans depend on Starship. According to its IPO prospectus, the document filed Wednesday listed Starship first among the risk factors facing the company. Interesting. Okay. So, a little bit more credit for you because I was like, ah, it's not that big a deal. I care more about the data centers. I mean, think about the optics. Some investors are super — you know, a little bit — it is interesting. Like clearly if SpaceX is listing Starship as the first risk factor, it means like, oh yeah, like Colossus, we're going to be able to do that. That's not that hard. We can build a big data center. They have possible — and there's less of a risk factor of like, we're not going to be able to monetize that. Because there are a group of investors out there that are saying like, 'Oh, token prices will collapse. GPU depreciation. It'll get blocked. There'll be a data center ban.' Like they won't be able to ramp up their terrestrial effort. God forbid like a catastrophic explosion — that would be dramatic. That would be potentially an omen that certain people would be looking into, but maybe — I have full faith. Yeah. Their execution record is unsurpassed, but this stuff is hard to get right, said Daniel Henson, portfolio manager at Neuberger Berman, who oversees a fund that owns SpaceX shares. He's got bags. This team in due course will execute on Starship and that will unlock tremendous value. They estimated in its IPO filing Wednesday that it has spent $15 billion developing Starship. Well, that makes a lot of sense why there's losses if you've been investing in this. Although it seems like the development costs are declining because in 2025 the expenses were $3 billion. So I don't know how many years, but if you imagine they've been working on Starship for five years, then three billion a year is sort of average. These efforts have weighed on SpaceX's bottom line with the space division reporting a $662 million loss. The interesting question that I have from the SpaceX IPO prospectus is — Dan Primack had that sort of hot take saying like, oh wow, a lot of people were expecting SpaceX to be a bigger business than what they showed in this prospectus. Obviously Elon is always looking to the future. Huge TAM, huge opportunity in space data centers, huge opportunity in terrestrial data centers. Like the revenue is going to grow. The $75 billion that we need is — he just added like $15 billion in ARR on a $20 billion company. Like the growth is there. Elon makes plays, but my question is as you trickle down, it's an AI company with a data center business with a telecom, and then the smallest section is the launch business. And so I'm wondering like, what does the P&L look like? What do the financial statements look like for Blue Origin? Because Blue Origin's been operating for I think a longer amount of time, maybe the same amount of time, 20 years. So I can't imagine that they're — if they're working on a new rocket, are they going to be — 2000. September 8th, 2000? 2000. So that's a 26-year story and they're building stuff that's trying to be in this class of rocket. And if Elon's spending three billion last year alone developing Starship, like what do the costs look like for Blue Origin? And then the revenues have to be so much.

Here's a question for you, John. There was a PR firm reached out trying to get an early SpaceX investor on the show. Yeah. And they invested in 2019. Does that count as being an early investor if the company was started in 2002? I think so. I think pre-Starlink counts as early investor because it's always early. It's always early. Well, also, early is often defined by the ROI on the investment. And so, if you got in in the decacorn rounds and now you're going out at a trillion, that's 100x. Yeah. And so, you're in a pretty good spot. I think — Sean McGuire at Sequoia, like it was a later bet. 2020, I think, was when he started building that position. But pre-Star — like Starlink really. Yeah, I think Starlink was starting to work or starting to be mapped out as a business, starting to be understood by the market as a potential profit center in the midst of a very difficult launch market that was, you know, not growing exponentially and there were good margins, but the launch business alone was never going to propel SpaceX to a trillion-dollar valuation. But you add two, three, four other businesses and all of a sudden you get there.

Other tests have ended in explosions, including one this past June that shook homes miles away. A failure early last year threatened commercial airline traffic in the Caribbean region, The Wall Street Journal has reported. The new version of Starship includes upgrades to the fuel transfer tube that moves propellant to the booster's 33 engines, redesigned so engines can simultaneously and more quickly begin firing. And I think they painted the engines black for this one. I saw Elon doing like a little side-by-side of the silver engines. He was saying that the black engines look better. I kind of agree, but I haven't looked too closely at the images. I don't know if it's — it's probably purely functional, but it is fun as these things get more design. You've seen those images of like the Raptor engine or the Merlin engine getting redesigned and then just removing parts, and then the final version looks pretty aesthetic. Like it looks pretty beautiful and the first version looks just like cables and wires and tubes everywhere. Looks very thrown together. There's something about efficiency that is also aligned with aesthetics sometimes. The engines, called Raptors, are expected to produce more thrust while weighing less, according to the company. Oh, that's good. Almost every part of Starship V3 is different from V2, Musk said in a social media post Sunday. The postponed test mission was set to deploy 20 Starlink simulators during the mission. I'm wondering how much a Starlink costs because why don't you just rip a real one? You know, you're spending three billion. Just put up the real ones on the case that it actually works. But I guess there's probably more like telemetry and simulation and data collection that you can do if you send up the simulator as opposed to the real thing. But I don't know, why not just rip a satellite? Remember — I mean, the ultimate simulation is just doing the thing. Yeah. Right. Well, Elon says we live in a simulation potentially. Exactly. So maybe that's what he's doing. What are the Starlink simulators really simulating? Who knows? DJ D'Souza. Yeah. What's going on with — sliding into Elon's DMs? Oh, yeah. In the bid to lead the IPO, with banks jockeying to get their name first on the deal for SpaceX's IPO, Solomon worked with staffers to message us directly. He's like, 'All right, guys, help me out here. How do I send a message on X?' according to people familiar with the matter. So he's putting in the work. And he says, 'Wow, this is how you work for the bag at Goldman Sachs.'

David Solomon also had a piece in The New York Times saying that the AI apocalypse was overrated. AI is a job creator, David Solomon says. He says, 'I'm the CEO of Goldman Sachs. The AI job apocalypse is overblown.' We'll see what he's saying in a couple months. Ken Griffin sort of flipped. Although Ken Griffin hasn't gone as far as to say that there's a job apocalypse. He was just saying that the models went from like slop to usable, but he hasn't actually opined on whether or not that will change Citadel's hiring plans or his view of America's — like cap on how much software — I was there. I was writing Visual Basic. It was brutal. I would have loved Codex and Claude Code. I would have been working for one minute a day instead of an hour a day. If you could have had Codex back then, just like a private — a private instance of it. Yes. You would be currently — time traveling back in time just to become the most productive intern of all time is elite. That's something you would actually do. Oh, it would be so much fun. Maybe that's what the future will hold. Intern simulator. You get to go into VR and simulate being the most elite intern with time travel technology.

Anyway, can we talk about the World's Fair? Yes, we need to talk about the World's Fair. Tyler, you gave this a read. What stuck out to you? Maybe let's read through some of it first. And quick update, we are rescheduling with Matt Grimm. Okay. Due to technical issues. Yes, we will get him on. He's been here in person. We know Matt Grimm. We love him. Always fun to catch up with him. It's possible that a third-layer SPV lead is holding him hostage saying, 'Give me direct access.' I think he might have been there, finished the Murf, and saw someone hustling an SPV and he just had to drop the mic and run off and tackle them. That's probably what happened.

Anyway, World's Fairs. Has someone tried to bring these back? This feels like a tech new media, 'We need a World's Fair' thing. Like I could see this being something that someone tries to revive. Although it'd probably be a lot of screens these days. But at the World's Fair of old, there were a lot of cool technologies, a lot of hardware, a lot of interesting exhibits. And we'll go through — I don't know how much you've studied the World's Fair in the past, but The Wall Street Journal has been reflecting on the United States of America at 250 years with a whole bunch of articles that look back and contextualize American history. And we'll go through this one on the World's Fair right now. So, among the groundbreaking innovations unveiled at US-hosted exhibitions were the telephone, the Ferris wheel, and electric lighting. Ferris really crushed it on the naming thing. Whoever — you know, Alexander Graham Bell invented the telephone. Thomas Edison invented lighting mostly. Those guys did not bake their name into the product nearly as well as Ferris because it's capital F and I believe Ferris is the one who invented the Ferris wheel. Anyway, before splashy Silicon Valley product launches and the annual Consumer Electronics Show, the world showcased its latest innovations at high-profile world's fairs. These exhibits, held every few years, beginning in London in 1851, gave countries a public platform to showcase new technologies, products, and ideas that were going to reshape daily life. For host countries and cities, the world's fairs or expositions were an opportunity to own the world's fair as a sharing project. The person that invented the Ferris wheel's full name was George Washington Gale Ferris Jr. That's pretty elite. George Washington — there's nothing stopping you from naming your child George Washington and then just slapping your own last name at the end. Yes. Yes. I mean — George Washington Cosgrove. My John Tyler was the 10th president. Oh, true. Yeah, there you go. That's my name. There we go.

Okay. So, many of the big expos were commemorative, says Paul Greenhalgh, a British historian and author of two books on world's fairs and expositions. But all of them ended up doing something else. They changed the shape of cities much like the Olympics today. World's fairs were large-scale planning efforts and economic investments as much as cultural events. Cities competed for the right to host them. Chicago was able to redefine its image from a gritty second city to a cultural and architectural hub after defeating New York in the bid for the 1893 exposition. Here's a look at the US-hosted World's Fairs that left the deepest marks. So, Philadelphia — starting in Philly — 1876, you're not going to believe what was shown, what was celebrated. We had two key technologies: telephones and ketchup. One was disrupted much faster than the other. Yep. I would say that the telephone has basically disappeared. The smartphone, text messaging — telephone's not doing well. It's on its last legs. Barbecue sauce really made — gained ground. No, ketchup is still — even ranch kind of went on a run, but really — mustard, ultimately it's a complement, not a substitute. There are very few people that say, 'Oh, I don't need the ketchup at all. Just the mustard. Just the mustard, please.' Because most people — I don't know. I do think mustard is better than ketchup. Than ketchup. Yeah. Interesting. Yeah. You can believe that. Yeah.

America's first officially sanctioned World's Fair, the Centennial International Exhibition, arrived when the country was still defining its place among industrial nations. Unlike earlier European fairs built around single monumental halls, the Philadelphia event spread across a vast park with hundreds of structures. It was sort of like the Coachella of its time. Among the most notable exhibits was the telephone. It was initially greeted with skepticism, but the potential of the cool contraption quickly drew attention. Once people realized what had been invented, it became clear how important it was. The fair also showcased Thomas Edison's automatic telegraph as well as machinery such as the typewriter. Imagine you just create the automatic telegraph and then you go to launch it and demo it and somebody invents the telephone. She's like, 'Really? Really, dude?' And consumer packaged goods such as Heinz tomato ketchup. It normalized the idea that private companies, innovators, and even governments could present their work side by side, a model that amplified the country's entrepreneurial culture. I love that.

Chicago, 1893. You're not going to believe it. You're not going to believe the two technologies that are — We got to call balls and strikes here. A little bit more underwhelming. Yeah. But I don't know. I think — Tell people what was — Tell people what was — two key inventions last time. Remember we created telephones and ketchup, right? And so fast-following this with the next fair we create the — No, no, no, no. It's not fast-following. 1876 — almost 20 years later. Almost 20 years later, they're like fast takeoff. We're so early. Telephones cooking. What do we got? The Ferris wheel and popcorn. Popcorn is pretty good. Ketchup before popcorn. Pretty good. Pretty good. But yeah, Ferris wheel. Ferris wheel. Imagine the hundreds of people that came there to demo their inventions. If you're in telephones, pivot to Ferris wheels. There's a whole industry. Oh yeah, Ferris wheel. That'll destroy the telephone. You got to get on the Ferris wheel. It's the new thing. Like it is total stag. It's sort of funny that like if you were to look back, if we had a World's Fair, you'd have like — creatine gummies and like AGI, right? Exactly. So in 1893 they had the Ferris wheel and popcorn and then you fast-forward to today it's like we found a way to put one of the micronutrients in steak into a candy-shaped supplement. And we need you to eat four or five of these a day forever. 2020 World's Fair. We have rockets that land. And we also have — Allbirds. Allbirds. What else came out? That was very silly. Bored Apes. We've invented Bored Apes. Skims shapewear. Shapewear and GLP-1s. Somewhat related. I don't know.

Buffalo, New York, 1901. Okay. An electric city. This seems pretty elite. This is pretty huge. 1901, though remembered mostly as the location where President William McKinley was shot. He died a week later. Buffalo's Pan-American Exposition was meant to showcase — He was shot. He was shot at the World's Fair. I had no idea. That's crazy. Fact check that. Buffalo's Pan-American Exposition was meant to showcase America's dominance and innovation in electrical power and infrastructure. The fair featured electric lighting and electric streetcars powered by hydroelectricity generated by Niagara Falls, reflecting a moment when electricity was beginning to move from novelty towards widespread utility. Okay. So this is 1901. 1901. And remember in 1876 we created the telephone. Yeah. And you fast-forward to the next event in St. Louis in 1904. You're not going to believe what we remember the 1904 World's Fair for. Computers. We created electricity and they're like, we got something better. What do we have? We now know this World's Fair as a snack food extravaganza. It's really just a gummy. It was sort of a Cambrian explosion of snacking. Yeah. I mean, this is the nature of our show. It's like Cerebras and Groq. Like this is actually what happens in America. The Louisiana Purchase Exposition was among the largest ever staged and nearly obliterated the city's finances. Oh no. Still, its pop culture influence endures. The exposition popularized a host of snack foods that would become staples of American life, including ice cream cones, peanut butter. Oh my, they went on an insane run right here. Ice cream cones, peanut butter, hot dogs, hamburgers, and cotton candy. Mhm. Elite. While the foods weren't all invented for the fair, their widespread availability there — like it's so funny. So people are coming there again to share their actual innovations and then snacks had just come online and so you can imagine like it just turns into a snack conference where everyone's just snacking like crazy. They're like, 'Wait, we have hot — you can go to a hot dog stand and then you can go get cotton candy.' Yeah. In one place. You got to be kidding me. I mean, I wouldn't be surprised. Visitors also encountered early examples of X-ray machines. So we invented X-ray vision, but history remembers it as a snack food extravaganza. Peanut butter — early flying machines and submarines came after — and automobiles. And automobiles and submarines. This is crazy. And barely remembered. No.

Fast forward 1962. Seattle. Space race. The space race. That's exciting. The World's Fair. By the mid-20th century, the World's Fairs had become games of one-upmanship. I like that. In geopolitical rivalry — feels like where we're at today. Seattle's Century 21 Exposition took place in the shadow of the Cold War. Following the Soviet Union's launch of Sputnik, the world's first artificial Earth satellite, the Soviets turned down an invitation. They turned it down. Turned it down. Turned it down. The federal government became deeply involved. There was a sense that America needed to demonstrate technological leadership. Mhm. The Space Needle was constructed for the fair. I did not know that. I recently saw the inside of the Space Needle because — wait, no, that's not the Space Needle. It's the Toronto version of the Space Needle because didn't Drake film a music video inside of it? He iced it out. Oh, he did. He made it ice blue because he's Ice Man or something. That's the name of the album. And inside you can see all of the broadcasting equipment and there's these huge radio dishes inside. It's very cool.

Anyway, fast forward just a few years, a couple of years, we got — you're not going to believe what they announced at this World's Fair. 1964. They invent — 1964, more than just under a hundred years since the telephone was invented, we invented punch cards. Punch cards, which feel like a kind of early version of enterprise software in some way. True. World's Fair, Flushing Meadows, New York City. Yes. Records. Visitors were dazzled by color television demonstrations and the Picturephone, an early video calling system. They had FaceTime. Yeah. FaceTime. This is sick. This is crazy. I can't believe — when I told Jordi 1964 they invented FaceTime, he was like, 'Did it use mirrors?' And he thought it was just a periscope. Just a mirror that you look at. Periscope? No, no, no. It's like a series of mirrors that bounce my —

...image across so you can be in the other room and you can see me and talk. No — no, they had full video calling. I guess AT&T's Picturephone, which added video to telephone calls.

Yeah, so it cost $16 for a three-minute call. That's $121 in today's money. That's expensive. Yeah. $40.

Token maxing. I wonder if there were CEOs at the time who were like, we need to be — if you're not spending $20 a year on your picture phone, you're not going to have...

I wonder if they were talking about Jevons paradox. Yes, probably. I don't know when the last officially sanctioned World's Fair in the US was — it was hosted in New Orleans in 1984 and it produced no truly defining technological debut. Stagnation theory undefeated. Brutal.

By that point the function of World's Fairs had largely migrated elsewhere — to museums, television, theme parks and technology conferences. As a result, the World's Fair concept gradually faded. 'Putting on an expo is a lost art,' the author says, 'but when America needed it, no one did it better.'

And Brenell, of the TBPN newsletter fame, shared that there was in fact an attempt to bring back the World's Fair, by Zack Dev. He wrote on Substack, StartupCities.com, in 2023. He talked to World's Fair co-founder Cam Whis — Walt Disney's Epcot, Dubai's Museum of the Future — why the World Fair died, and why big in-person events matter more than ever.

So, to push hard, you could bring something like this back, but I feel like it would just turn into a tech conference, you know.

I don't know if that would work. Anyway, um, what else stuck out to you, Tyler? There was another article you put in here from The New York Times about the picture phone.

Yeah. People aren't really...

Well, you could just say that I think of you as an inventor of bits.

A lot of — I mean, even also just the way inventions take place is so disaggregated. Like, we were doing this during the GLP-1 boom, we were trying to trace through, like, who are the key players. This is the take on, like, we don't have ticker tape parades. Like, why is there not one person? Doudna got credit for CRISPR. There's a book Walter Isaacson wrote about her, I think it's called The Inventor, maybe, I don't know. Doudna, she won the Nobel Prize. Um, but...

The internet is the new World's Fair, right? Or X or any of these social platforms. It would have made sense in the 20th century, in the 19th century too, to create something and think, okay, I have to travel somewhere to a place where I can find a big audience of people that are going to be interested in new inventions. And now people create something and it goes viral on the internet. So there's a film called The Inventor and it's about Elizabeth Holmes, right? So that just shows you how much the idea of an inventor — it's like mocked now.

People don't have respect for an inventor. It's the scammy grifter. Who's the inventor? It truly, truly is. Well, we have someone who ships — we have Dan Shipper in the waiting room.

Let's bring him into the TBPN Ultradome. Let's catch up with Dan Shipper. How you doing?

You're a founder. You're a shipper. You're an author, a writer. Do you think of yourself...

Yeah, you do. You're an elite ship athlete. It's great to see you. It's been too long. Do you think of yourself as an inventor?

Even though — yeah, I don't know. Inventions can be patented. People that invent software, you should give people credit for those things, but we've moved away from that term for better or worse. Anyway, it's been — maybe has it been like four months, maybe more since we caught up. It's been a while.

The world's changing. Yeah, give us the update on all things Every and Dan Shipper.

I saw an interesting benchmark here where someone outside of technology was lamenting AI progress because the tricky prompt they had was come up with a Pokemon that ends in the two letters 'er'. Did you see this? And the base models, the cheapest things, the older models sort of get confused by it because of the tokenization. And it's sort of like a new version of how many Rs are in 'strawberry'. But if you hit a coding model, like you can just watch the traces of Claude Code or Codex and it's like, go to Pokemon Wiki, download them all, put them into a CSV, write the Python that checks the actual last two characters of the Pokemon name, and it just nails it perfectly, and it's traceable, and so you get the perfect answer. And I think people haven't realized that that's what's possible and it doesn't all need to be memorization anymore. That's like the big shift.

Yeah. I mean, it's a big shift because you're going from everyday knowledge work being good at writing documents and good at collecting information — which are valuable things — but you've never worked with someone in an organization who was a superstar where all they did was gather information and collect reports and things like that, right? It's like, okay, what can you do with that? What kind of actions can you take? And yeah, it's been — even with how close we are to the action, you know, we weren't doing the show in the GPT-3 days, but I still find myself operating day-to-day thinking like, oh, I should reach out to that person to see if they can help with this thing, and then realizing like, oh, I can just ask Codex to do that, or I can get a model to do this. And it's a fundamental shift and I'm still not routing enough tasks through the models, or at least as many as I could.

The other thing I wanted to get your update on is how you feel like memory is evolving. I've had some pretty pleasant experiences recently where I'll go to ChatGPT and ask for something and it's actually able to pull in a bunch of really relevant context and memories in a way that is helpful and not just kind of distracting or annoying. And I specifically wanted to get your update because, like, call it 18 months ago — probably a year ago, 18 months ago — I think everyone was saying memory, memory is the moat, like memory, and then people just kind of stopped talking about it for six or nine months or something like that. But it feels like it might be having a moment too.

Yeah, it's interesting because it's very popular now to say AI in the layoff announcement, but it's also very popular to say, and our business is better than it's ever been. What do you guys think about that? I just feel like that's such a weird strategy or weird meme to be going around.

So the CEO of Standard Chartered came out and said AI is working so well we're going to reduce our workforce, and he used a really spicy phrase. He said low-value human work, and we're going to replace it with capital that we're investing. It was a very quantitative analysis of something that's deeply human. But I dug into his actual projection and he said 15% headcount reduction by 2030, which is like — I would assume that attrition would get you there. Like, is this the most retentive company in history? Like, I feel like 4% of the workforce probably turns over annually anyway. You can just slow your hiring plan and get there if you want to shrink. And that's what a lot of companies have done in previous eras where there's been hiccups or disconnects between valuations. There were some big private companies that during the COVID era or ZERP era just sort of froze hiring and grew into their valuations because the business model was sound. They didn't need to hire as much. They refocused, raised the hiring standard, allowed attrition to play its part, and they got to become leaner and meaner organizations without actually doing a layoff — which at that time would have been optically very bad, because if you were doing a layoff during ZERP or COVID it was like an indictment of you. It was like, oh, you're a victim of COVID or you're a victim of ZERP. Your business doesn't work in a high interest rate environment. Your business doesn't work in a more remote world. Whereas now the vibe is, oh, if you're doing a layoff, that's a sign that your business is ready for AI, ready for automation, ready for more efficiency. And so I think people are sort of playing both sides.

How are you thinking about — this has been talked about plenty at this point, but I'm curious in the context of 'After Automation' — this difference between like a job and like a task, right? So AI right now very clearly can do a lot of tasks, but it cannot do very many jobs. Or really any jobs. Or like, there's a subsection of jobs that AI can do well 90% of the time, but even that is not enough to replace the role or function entirely.

Excellent. What do you think about the idea of the YouTubification of software? Like, Hollywood has been on a decline as content creation has been commoditized and democratized, but you've seen many more small creators pop up, many more lifestyle businesses, many more small businesses get off the ground. And it feels like that is one possible direction that this goes — more niche software. You used to need to raise a whole Series A from a venture capital firm to rack servers. Then with AWS, you needed to go through YC, be technical, hire a couple of engineers. Now a few people can sort of get something off the ground. Are you actually seeing any movement towards that, do you think?

It is a marriage violation. Yeah. The only thing I think about is oftentimes you're buying software for a certain group of people to sort of manage work, and as certain workflows become agents, then — like another way to put it is, a lot of people outside of tech who have been vibe coding are sharing apps with me that they've built that the LLMs can just do natively pretty well already. And so that's kind of the potential bear case for this super, super long tail of software: as agents get more competent and people learn how to — they're sort of unhobbled — there's a lot of this long tail that can just be like a thread, effectively.

Oh, you're saying you don't need an app, you just need to talk to people.

You could vibe code bench press. Yeah, at distribution or the customer could just wind up going to any...

Love it. Do you think your agents are planning a surprise birthday party for you?

Happy birthday. Do we hit the gong?

Awesome. What can people expect from Every over the next — call it a month, because that's like a year in AI years.

Amazing.

Thank you so much for coming on the show. We'll talk to you soon. Goodbye.

Up next we have James Rogers from Apeel Sciences. He is the founder, here with us live in the TBPN Ultradome. Maybe move these objects out of the way so we give you a proper entrance. Thank you so much for coming on down. How you doing?

Please introduce yourself for everyone who might not be familiar. Introduce yourself and we'll go through the story of the company and talk about the news.

Wait, yeah, why Santa Barbara? I remember finding this out and it being an interesting tidbit. There's some cool companies up there — Sonos, and isn't Dyson up there a little bit or something?

Okay. What does that do? You painted...

Okay. Where did that technology go? Is that still —

Power's expensive now though. Maybe the economics work in the data center era. You just paint the data center.

What was your reaction to touring UCSB for the first time? I'm a gaucho as well.

Because from the air — my favorite thing is if you show someone a picture of the university from the air, it looks like it should be a Four Seasons or something. A university has no business being that close to an incredible beach.

It doesn't make any sense.

So you study material science. Where did the original idea for Apeel come from?

So the initial product — are you in the lab? Are you on a pitch deck? Are you raising money and then going and doing R&D?

That is true. That is true.

Just put those words together.

Yeah. So what was the specific founder journey? You have this idea, you tell your mom about it. She's like, 'Good, that's nice, sweetie, but you don't know anything about fruits and vegetables.' And then what do you do next?

So, grant money early on. When does the business...

New Venture Competition?

At UCSB. Yeah, I remember NVC. But wasn't that like — didn't you get like 25 grand or something?

No, I remember this program and yeah, it was really like venture funding for ants.

My first company, at least twice the size. 17k, first company for me.

It's enough if you're eating ramen and just paying rent.

No way. That is crazy. That's probably the most enterprise value created from an incorporation on DP.

Who knows? You'll probably get some — Del Playa is just a very notorious street. It's the street that runs in Isla Vista, which is where the university is. It's a street that runs along basically as close to the water as possible. You want to be on the — you're usually surfing if you're on that street.

Not necessarily.

Maybe a little bit. Maybe a little bit.

Good. No, that's actually good. That means you're way more locked in. Basically, the middle zone is a nightmare. I don't think any enterprise value is being created there. You're farthest from the university, but you're closer to nature. It's a little bit quieter there, but you go two blocks down, negative enterprise value creation.

Yeah. Okay. So you kind of have this insight. You have a little grant funding. Are you able to make MVPs with that grant?

How hard could it be? You just put some stuff on an apple and leave it out. And I'm joking when I say that.

No, no, but kind of — I mean, that was how simple it was supposed to be.

And so would you effectively go buy two apples, hopefully from the same orchard, put...

No way.

Wow. That's crazy. That's globalized. We'd be growing them in the summer.

So yeah, talk about the early business model — the farm side, the table side, distributors, supermarkets. There's a lot of different players in the supply chain that might be interested in the business. How did you think about cracking it?

The garbage can is my best customer.

Okay.

Did you ever get a visit from Big Fruit coming in saying, hey buddy?

They were — you know, you guys probably have some of their masks around here. I've seen it. Yeah, under different clothing.

What does that mean?

Okay, interesting.

So you think they were basically lying to you?

Oh, interesting.

Okay.

Because of the joke with the avocado. Not now, not now, not now, now too late.

Okay.

Dude, my UCSB days, avocado was such an extreme luxury that the joke of like, 'not now, not now, not now, now it's bad' or whatever, that was so real. I was like monitoring my avocados, being like, I cannot let this go.

Yeah.

They don't even know. Yeah.

So retailers would actually acquire produce from a distributor and then apply Apeel to the product.

Okay.

Yeah.

Yeah.

Sure.

And so, what year is this when this starts working in Europe?

Oh, wow.

And you had already raised at that point like hundreds of millions of dollars, right?

That was the world supply, right? Bet it all.

So you raised some money. What did it look like actually scaling up your supply chain?

That's not going to work.

Distillation.

Yeah.

By this point, did your mom think that you had figured out fruit?

Yeah. I mean, I've spent basically my whole life in California and so always had access to farmers markets. And sometimes the farmers markets are selling product that isn't actually, the farmer is not present and maybe not even in the state. But oftentimes it is hyper local. So you get super spoiled living here. And one thing I've come to realize is if you don't know the person growing your food, they don't have a lot of incentive to do the right thing for you.

Mhm.

You know, it's just people sell food based on price. And if I have a lemon that's coated in pesticides and a lemon that's not coated in pesticides, you can't tell that by looking at it. And this one's cheaper.

The one coated in pesticides is cheaper because they have less rot. So you buy the cheaper one and the person who's doing the right thing doesn't get rewarded for it. So there's just this horrible incentive in these supply chains to cut corners and do stuff that is horrifying.

What did you learn about organic as a certification and the issues of that? We had Bryan Johnson on the show recently and he was saying he oftentimes will assume that organic food is going to be more contaminated and worse for you than the non-organic just because there's all these workarounds.

Oh, interesting.

What is an inorganic apple? Just like computer chips and aluminum or something? Aren't they all organic? They are literally organisms, right?

Definitionally.

But I mean, that's ultimately part of the problem is that you were sort of attacked with an aggressive marketing campaign against your product, right? And it kind of cuts both ways. Take us through that part of the journey. Did it start like a low rumble and then turn into like a firestorm?

Yeah. So, set the table for us because you've raised some money. The business is working. The product is working. You're in enough stores at this point. Your mom is buying hundreds of...

Pulling off the stickers. We got distribution. Not like she has to go to some special store.

Boom.

Yeah.

It was a similar name, was that...

Exact same name.

So it's not even a shared ingredient.

Okay.