Trump-Xi Summit, Inflation Fears, Space Data Centers | Max Levchin, Delian Asparouhov, Richard Socher, Brandon Hill, Nate Tepper, Joubin Mirzadegan, Roman Chernin

Summary

---

TBPN Podcast Summary

Episode: Trump-Xi Summit, Inflation Fears, Space Data Centers (May 13, 2026)
Hosts: John Coogan & Jordi Hays
Guests: Max Levchin (Affirm), Delian Asparouhov (Varda), Richard Socher (Recursive/You.com), Brandon Hill (Vori), Nate Tepper (True Short), Jubin Murzadegan (Roadrunner), Roman Chernin (Nebius)

---

Overview

This packed TBPN episode traverses the collision of tech, geopolitics, and the new economic realities. The show unpacks the high-stakes Trump-Xi summit in Beijing with major business and AI implications, tackles inflation trends, explores orbital and oceanic data center dreams, and features a gauntlet of top founders unveiling fresh milestones in fintech, pharma-in-space, AI R&D, grocery ops, short-form film, enterprise revenue infrastructure, and hyperscale AI cloud. Notable moments include insights into US-China AI arms dynamics, the infrastructure/funding crunch for chips and energy, debates about the future of humanoid robots, and the challenges of building durable startups amidst rapidly shifting technical and capital environments.

---

Table of Contents

1. Trump-Xi Jinping Summit: Stakes & Tech Implications
2. AI, Anthropic, and China’s Catch-Up Dynamics
3. Orbital, Blimp, and Ocean Data Centers – The Next Compute Frontier
4. Inflation Worries & The Trough Economy
5. Interview Highlights
   • Max Levchin on Specialization & AI’s Impact
   • Delian Asparouhov on Manufacturing Pharma in Space (Varda)
   • Richard Socher on Recursive Self-Improving AI
   • Brandon Hill on Grocery OS (Vori)
   • Nate Tepper on Vertical AI-First Short Films (True Short)
   • Jubin Murzadogan on Revenue Infrastructure (Roadrunner)
   • Roman Chernin on AI Cloud Growth (Nebius)
6. Memorable Quotes & Moments
7. Timestamps for Key Segments

---

<a name="trump-xi"></a>

1. Trump-Xi Jinping Summit: Stakes & Tech Implications

• Cast of Characters: Trump and Xi’s “stock summit” in Beijing includes Elon Musk, Tim Cook, Jensen Huang (Nvidia), Kelly Ortberg (Boeing), David Solomon (Goldman Sachs), Stephen Schwarzman (Blackstone), Larry Fink (BlackRock), Jane Fraser (Citi), Dina Powell McCormick (Meta).
• Why Jensen? Speculation about Nvidia CEO Jensen Huang’s strategic import, his late addition to the trip, and the symbolism of US chip tension with China.
  • “[Nvidia] is maybe the most tension between trade deals with China.” (Host, 02:59)
• Red Carpet Arrival: Trump, Musk, and Jensen land on Air Force One to fanfare; hosts riff on the symbolism of manufacturing might (planes/chips), personal branding (Jensen’s “uniform”), and how executives frame their roles.
• Summit Stakes (Wall Street Journal analysis):
  • US aims for “stability,” seeking truce on tariffs and rare earths—not fundamental change.
  • Xi’s priorities: AI, rare earths, and Taiwan. China aims for access to US chips/models and pushes for US to tone down support for Taiwan.
  • AI “arms control” is largely performative (“don’t expect much”), US dominance in AI models and compute seen as best deterrent.
  • Noticeable absence: Heads of leading US AI labs (Google, OpenAI, Anthropic) aren’t on the trip—reflects a gap between Silicon Valley priorities and DC high-level geopolitics.
  • “You’re not getting answers to superintelligence or post-AGI questions at this summit—it’s mostly about supply chains and semiconductors.” (Host, 15:36)

---

<a name="ai-anthropic-china"></a>

2. AI, Anthropic, and China’s Catch-Up Dynamics

• Claude Mythos Goes Abroad: Anthropic opens its powerful Mythos model to Japan’s “three megabanks,” with approval from US Treasury. National security concerns are top of mind, as the model can find software vulnerabilities at scale.
  • “The Secretary of Treasury is personally vetting and approving each company that gets access to the most advanced US AI model… Meanwhile, Jensen Huang is flying on Air Force One with Trump to Beijing to sell AI chips.” – Host (21:55)
• China's AI Progress: Dario Amodei (Anthropic) estimates China’s open-source community is ~6 months behind Mythos. What would China’s “Mythos moment” look like? Discussion contrasts US private-sector-driven, semi-democratic oversight with the CCP's direct and often abrupt assertiveness.
  • “What happens when the CCP starts to have more control over a company?” (Tyler, 26:44)
• Energy & Compute Control: In the West, regulation often centers on software and audits. In China, control is more about gating chip/compute supply at a physical level.

---

<a name="orbital-data-centers"></a>

3. Orbital, Blimp, and Ocean Data Centers – The Next Compute Frontier

• SpaceX & Google’s Orbital Data Centers: Major talks to launch “space data centers.” A million satellite filing hints at massive ambition; debate on technical challenges (cooling, radiators, power), insurance, and regulatory weirdness.
  • “Space is free, the land is free. Cooling, potentially free energy. Little bit tricky down there.” (Host, 30:00)
• Other Models: Amazon/Blue Origin in the game. Sergey Brin's blimp company is floated as an underappreciated angle ("blimp data centers"), as is ocean-based ("Atlantis" undersea servers leveraging cooling and space in the Pacific).

---

<a name="inflation-worries"></a>

4. Inflation Worries & The Trough Economy

• Inflation Hotter Than Expected: Producer prices up 6% YoY vs. 4.8% estimates. Energy price surges are driving market anxiety about stagflation (stagnant growth + inflation).
  • “The real economy is only growing at 0.1% while suffering from 2.7% inflation—a recipe for stagflation.” (Host, 60:05)
• Trickle-Down to Tech: Debate touches on the concentration of GDP/economic growth in data center/AI CapEx, while “real life” remains tepid.

---

<a name="interviews"></a>

5. Interview Highlights

<a name="levchin"></a>

A. Max Levchin (Affirm) – Depth Wins in an AI World

• On Career Choices:
  • “If you out-depth your competitors, they just can’t beat you… Depth is a profound strength.” (Levchin, 63:29)
• AI and Specialization:
  • Advocates for extremely deep expertise, especially in computer science, even as AI tools create more generalists.
  • “The unattainable mount olympus of value: being the expert the AIs want to learn from. That’s where the value is.” (65:39)
• Affirm’s Performance:
  • Recent investor forum: “We’re going to compound at 25% every year and move profitability floor up… It’s like a flywheel, the network spins faster and faster.” (68:15)
• AI Internals:
  • “Shipping code is night-and-day faster. AI writes 75% of code at Affirm now… and it’s only getting more productive.” (76:54)
  • On “token-maxing” for AI: “We monitor ROI for every team’s token spend, weekly. It’s already accretive but not ‘infinite budget.’” (79:10)
• Evaluating AI Products for the Enterprise:
  • “If you can define the evaluation criteria for the tool, it’s worth piloting. Otherwise it’s slop and you’re just being sold a story.” (85:06)

---

<a name="asparouhov"></a>

B. Delian Asparouhov (Varda) – Manufacturing Pharma in Space

• Announcement:
  • United Therapeutics is partnering with Varda to make the first physical pharma products in low Earth orbit. “First time a $25B public company is going to make real products in space.” (91:24)
• Why Space?
  • Microgravity enables new methods of drug crystallization—e.g., “changing administration from IV to subcutaneous, or oral to inhalable.” (93:43)
  • “One kilo of Ozempic retails for $1 million—a dose is mostly powder, so space manufacturing can be economical for high-value drugs.” (104:01)
• Process & Roadmap:
  • Runs through tech stack: two-part satellite/pod, bioreactor control, temperature/vibration issues.
  • Future: From disposable pods to reusable “space planes,” then possibly to a true permanent space manufacturing facility.
• Why Not Space Data Centers?:
  • “I don’t want to fight SpaceX in their back yard—pharma TAM is massive, more durable, and less competitive than AI GPU hosting.” (117:23)
  • “Once AGI arrives everyone will focus on longevity and biologics—happy to build the infrastructure ahead of that wave.” (120:30)

---

<a name="socher"></a>

C. Richard Socher (Recursive/You.com) – Recursive Self-Improving AI

• Company Announced: Out of stealth with $650M raised (valuation $4B, led by GV and others, Nvidia/AMD strategic). Focused on recursive self-improving superintelligence to automate knowledge discovery. (123:49)
• Team Pedigree:
  • Co-founders include creators of RAG, Genie3, Vision Transformer, RL at Meta, Codex at OpenAI, and more.
• Strategy:
  • “Laser focus on enabling AI to automate the scientific method… once we have automated research, everything else follows.” (123:04)
• On Productization:
  • Might accelerate public offerings and APIs given “breakthroughs already happening.” (130:59)

---

<a name="hill"></a>

D. Brandon Hill (Vori) – Grocery OS

• Background: Third-generation grocer. Vori is “the self-driving operating system for supermarkets.” (139:10)
• Market Gap:
  • “Grocery is still paper and pencil but accounts for $1.5 trillion of the economy.”
• What’s Broken:
  • Manual pricing is the norm; Vori “does in one click what used to take twelve steps with a highlighter and bifocals.”
  • “We find five points of margin left on the table, automate it, and give it back to the stores.” (144:13)
• Tech Stack: E-ink labels (like Walmart’s) rolled out to independents, deep integrations with ERP/POS, serving the “other 75%” not served by Walmart/Amazon.
• Funding: $22M raised, $500M payments processed. (145:05)

---

<a name="tepper"></a>

E. Nate Tepper (True Short) – AI-Generated Vertical Films

• Idea:
  • Inspired by “micro-drama” vertical episodic soaps in China (“making more money than box office… 50 million a month in revenue”). (146:47)
  • Short, vertical, mobile-optimized, AI-assisted film production and streaming.
• Company: Vertically integrated studio and platform (content + app + monetization).
  • “Full stack creative company, vertically integrated.” (146:25)
• Execution:
  • True Crime as a wedge vertical is already #1 with $3M annualized rev.
  • “Creative pods” combine showrunner, AI filmmaker, and editor to produce a series weekly.
• Distribution:
  • “Growth hack” mentality; owns multiple TikTok accounts and leverages self-clipping, but also runs significant paid acquisition.
  • Eyes on genre-expansion and possibly enabling “the next Star Wars from a bedroom.”

---

<a name="murzadegan"></a>

F. Jubin Murzadegan (Roadrunner) – AI-Native Revenue Infrastructure

• Problem:
  • CPQ (Configure, Price, Quote) and revenue systems are “the most broken” software processes in the enterprise.
  • “Every 10 years there’s a new CPQ vendor because every 10 years, pricing models change (SaaS > AI consumption).” (160:54)
• AI’s Edge:
  • “Models are good at reasoning with the rules/policies for quotes, so you need an agent architecture + deterministic system + flexible pricing model.” (163:47)
• Sales Complexities:
  • Many stakeholders (CIO, CRO, RevOps, Finance, Sales)—hard both to build and sell.
• Fundraise: $27M total, backed by KP and Founders Fund. (165:35)

---

<a name="chernin"></a>

G. Roman Chernin (Nebius) – The AI Specialized Cloud

• Nebius Origin:
  • Spun from Yandex; “less than 2 years old but posting 841% YoY revenue growth.” (166:23)
  • Yandex alumni also spawned Clickhouse; “the talent bench at Yandex is genuinely insane.” (168:10)
• Company:
  • Full-stack, vertically integrated AI-specialized cloud (“new cloud” is too broad a term).
  • “Not just data centers and hardware renting—a true cloud experience with flexibility, developer tools, and fast time to value.” (168:17)
• Trends:
  • Lagging-edge inferences become cost-driven: “Open source catches up within 3-6 months, and enterprises will switch for 4x cost savings if they can fine-tune models.” (170:55)
• Biggest Challenges:
  • Physical bottlenecks; getting gigawatts of power and GPUs online is the hardest operational challenge.
  • Financing is now as big a bottleneck as semiconductors.
• Key Insight:
  • “This is a delivery business… it’s boring, relentless execution across supply chain, infra, software, post-sales support.” (177:00)
• Strategy:
  • Acquisitions are about moving even faster (“every 3 months you can lose a huge piece of the pie”).
  • Inference optimization stack (model and systems) crucial to profitability and customer expansion.

---

<a name="quotes"></a>

6. Notable Quotes & Memorable Moments

• On Political Brand-building:
  • “Trump… didn’t he famously call himself a stable genius?” (12:33)
• On Joker Moves at the Summit:
  • “Who wouldn’t want to arrive on the cool plane?” (03:09)
• On the Absence of US Frontier AI Labs:
  • “You’re not getting answers to superintelligence or post-AGI at this summit.” (15:55)
• On China’s AI Trajectory:
  • “Something is going to happen there where China’s going to have their own Mythos moment… going to be dramatic.” (24:58)
• On Blimp Data Centers:
  • “There’s no NIMBYism in the middle of the Pacific Ocean with blimp data centers floating around…” (29:18)
• On All-LP Overlaps:
  • “Everyone owns everything, everyone has diversified into every company at this point.” (32:17)
• General Catalyst Ad Szn:
  • “Marc Andreessen quote-tweeted it, said, ‘Stay tuned for our upcoming ad campaign: we’re the VC who doesn’t sneer at your idea.’ Getting spicy!” (53:36)
• Jensen’s Leather Brand:
  • “The clothing industry doesn’t want you to know this, but a uniform is incredibly underrated.” (06:17)
• Max Levchin on Depth:
  • “If you out-depth your competitors, they just can’t beat you… You are the one the AIs want to learn from.” (63:29, 65:39)
• Delian Asparouhov on Space Data Centers:
  • “I don’t want to fight SpaceX in their back yard… Pharma TAM is massive, more durable, less competitive.” (117:23)
• Brandon Hill on Inventory/Pricing:
  • “We find five points of margin left on the table, automate it, and give it back to the stores.” (144:13)

---

<a name="timestamps"></a>

7. Timestamps for Key Segments

• Trump-Xi Summit cast, who’s really at the table, business/media analysis: [00:30–18:00]
• Anthropic, China’s access, Mythos, supply chain diplomatic angles: [20:26–27:56]
• Orbital & ocean/bottom-of-the-sea data centers, Starlink/SpaceX, new IPO math: [28:41–38:21]
• Inflation segment—stagflation worries, energy, AI-driven stag economy: [57:21–62:01]
• Max Levchin Interview (Depth, Affirm’s business, AI codegen in orgs): [61:43–87:28]
• Delian Asparouhov & Varda, space pharma, manufacturing in microgravity: [91:06–121:05]
• Richard Socher—Recursive, automating AI R&D, model autonomous improvement: [121:06–132:35]
• Brandon Hill—Grocery OS, Vori, pricing infratech: [138:27–145:47]
• Nate Tepper—True Short, vertical drama, the new mobile Hollywood: [146:15–158:59]
• Jubin Murzadogan—Roadrunner, AI CPQ/revenue stack: [159:01–166:07]
• Roman Chernin—Nebius, hyperscale cloud, infra/finance/AI bottlenecks: [166:08–181:39]
• Rumors, Ads, VC Branding Skirmishes, wrap-up: [183:19–185:53]

---

Additional Memorable Segments:

• Meta Uniform Discussion (Jensen, Jobs, branding via clothing): [05:59–08:08]
• Unitree mecha/robot reveals, police-use-case dystopia, robots “breaking down brick walls”: [44:15–48:20]
• Office “whispering to AI”, dictation trend, becoming “noisy call centers”: [132:41–136:14]
• Robotics “live stream” debates—teleoperation or real autonomy?: [136:29–138:29]
• Fintech memes: “Are you the final boss of coffee?”—Levchin’s smelting rabbit hole: [62:01–63:12]

---

This summary covers all key discussion threads, guest deep dives, and notable moments, structured for easy navigation and comprehension by those who didn’t catch the live episode. All quotes/time attributions preserve the original speakers and tone.

Transcript

[Host 1]

You're watching TVPN.

[Host 2]

Today is Wednesday, May 13, 2026. We are live from the temple of technology, the fortress of finance, the capital of capital. Bunch of news going on today. Huge show. Max Levchin's coming on the show. Delian's coming in person from Varda. Recursive is coming out of stealth. Brandon Hill from Vori, Nate Tepper from True Short. Very excited to talk about vertical. The trough. The trough is cooking. Some of the trough is cooking.

[Host 1]

Trough economy.

[Host 2]

Juban Murzadogan from Roadrunner and Roman Chernin from Nebias. Absolutely destroying earnings. Very excited to talk to the co founder and chief business officer over at Nebias. But the big news this week, and I'm sure what everyone will be focused on is of course on the COVID of the Wall Street Journal. That's the Trump Xi Jinping stock summit that's happening in Beijing. Trump and Xi Jinping vie for wins at summit. They're duking it out. And Trump brought, they both brought a big crew with them.

[Host 1]

A team.

[Host 2]

Yeah. Trump brought geopolitics, a couple team sport, couple political people that we won't even care about. But the business leaders are what is particularly interesting. So we got Elon Musk from Tesla, Tim Cook from Apple, Jensen from Nvidia, Kelly Ortberg from Boeing, Little flex there because Boeing has been able to maintain the lead in manufacturing large planes for a very long time. DJI has not been able to scale up and make a 747. What's going on over there? David Solomon from Goldman Sachs, Stephen Schwartzman from Blackstone, Larry Fink from BlackRock, Jane Fraser from Citi, and Dina Powell McCormick from Meta. So we should read through the stakes of the Trump Xi Jinping summit in the Wall Street Journal from the Wall Street.

[Host 1]

Before we get into that editorial board, let's get into the more important stuff, which is do you think Jensen was always coming on the trip? Right. There's been some speculation. He got on the plane in Alaska.

[Host 2]

Oh, yeah.

[Host 1]

Was that just the most convenient place for him to get on or was he a late addition?

[Host 2]

I don't know. Why would he have been a late addition? I don't know. It seems like.

[Host 1]

Well, because there was reporting earlier this

[Host 2]

week he runs the largest company in America.

[Host 1]

I know. But there have been statements made from what I saw that he wasn't going.

[Host 2]

I don't know.

[Tyler]

Yes. Well, so Trump had a truth social post that said CNBC incorrectly reported that the great Jensen Huang of Nvidia was not invited. But then he corrected it he said he did invite him personally.

[Nate Tepper]

I don't know.

[Host 2]

These schedules move around. It's gotta be really, really difficult if you're the leader of the biggest company in the world to move things around and commit to this.

[Host 1]

I know, but what's more important for Jensen and his objectives?

[Host 2]

No, of course it's critical importance, but at the same time there might have just been other moving pieces of like, okay, when can he go? Where will he arrive from, where will he get picked up from?

[Max Levchin]

I don't know.

[Host 2]

These things always shift. I don't have a good thesis. I mean other than the fact that that is the company that has maybe the most tension between the trade deals with China. Like Apple obviously does a ton of business in China, but has been much less of a focus in the chip sanctions. And so having Jensen there a little bit more of a discussion point. What do you think?

[Tyler]

I mean, I think on China like the Nvidia stuff seems like pretty like fine now, but. Yeah, but Meta I think has like some conflict, right. Because I mean Manus. Yeah, maybe it's like not that big of an acquisition on like market cap

[Host 2]

wise, but yeah, yeah, blocking it. A lot of these things, you know, in the, in the context of Silicon Valley, it's like a $2 billion deal or a company that's doing 40 billion in revenue and then you get over there and the stakes are like, you know, a trillion dollar oil market or like all the rare earths in the world. And it's your, you know, geopolitical conflict, which is obviously a much bigger deal. And so I don't know, we can speculate it on more. There might be more reporting that comes out about Jensen's goals there, what actually happened, how he found his way onto Air Force One. But that was confirmed by Elon Musk. Right. He said that Elon Jensen and Trump were all on Air Force One together, flying over. I certainly wouldn't want to get there. I would want to arrive on the cool plane. Can we play the video of the arrival? Trump arrives in Beijing with Elon Musk and Jensen in tow. I guess the rest of the business leaders came separately, but here is their arrival with. Are those cheerleaders or something? I don't know. What they seem to be waving? Oh, flags. They're waving flags. That makes sense. Well, there they go, walking down the red carpet. Air Force One is truly stately. It's so disappointing that the economics of the airline industry didn't just get bigger and bigger. And so you find yourself very rarely on 747s, unless you're flying across the world, because they do have such a presence and they're spacious inside, but we have been, you know, cramped down into the 7:30.

[Host 1]

Are there videos, pretty much everything where they give a proper tour of air force one or is it kind of locked down?

[Host 2]

I think they've done tours. I mean, there isn't. There are several air force ones, I believe, that are on display at presidential libraries, and so you can go and tour them. There might be one at the Smithsonian. I'm not exactly sure, but I'm pretty sure you can go walk around.

[Host 1]

Old ones take note.

[Host 2]

I wouldn't be surprised if they keep the current spec, like, somewhat secretive. There we go. This is very grainy footage. It looks like. Oh, maybe it's just the middle of the.

[Host 1]

It reminds me of Dune.

[Host 2]

Dune Arrakis. That's Taiwan, though. Anyway, let's go through Jensen's pack list. He was having trouble thinking of what was he going to bring. Of course, he landed on just a ton of leather jackets.

[Host 1]

The back and even the sneakers, I feel like, are pretty iconic.

[Host 2]

Yeah. Like the uniform is really starting to compound where it used to just be the jacket. Soon we will know the exact brand and specs of the pants as well as the shoes, and the whole Jensen uniform will be out in full force.

[Host 1]

Yeah. The clothing industry doesn't want you to know this, but wearing just getting a few pieces of clothes that you like and then wearing them until they fall apart is incredibly underrated. Not saying that's what Jensen's doing, but it's pretty funny. I tend to wear almost the exact same thing every day. And so sometimes people on the team will come in and be like, oh, we're matching today. Like, every day you dress like this, you will match with me.

[Host 2]

Yeah, that's right. That's right. There's obviously a personal brand benefit to adopting a uniform. We've seen it with Jensen, Steve Jobs at the black turtleneck, Palmer Luckey with the. With the Hawaiian shirt. Big news from Anduril today. New round series. Are they running out of letters? Series H. H Series H51.61 billion, something like that. Fantastic progress. But you adopt the uniform, you become more iconic, you become more recognizable outside of your face, and that builds your personal brand, your lore, and it helps you sort of differentiate and stand out. Right. The flip side, the thing you got to watch out for is that if you are always dressed exactly the same and you never switch your wardrobe, you, clips of you saying something a decade ago can go Viral. And if you're not, if you're aging gracefully, it will look like you said them just yesterday. And that can be a big problem if, say you're an AI CEO who a decade ago is reading some sci fi and thinking like, man, there's something to this Terminator thing. And then maybe in the past decade you've done a lot of work. You've investigated the systems. You have a more informed position and opinion. Well, the old video of you still looks like you today. And so there might be something to leveling up the wardrobe and changing the wardrobe over time that actually creates more distance between. Oh, well, clearly that was.

[Host 1]

So this is an indirect way of saying you want Jensen to be in a leather sort of like, you know, sort of trench coat eventually, like flowing robes.

[Host 2]

Well, the easiest thing to do is go bald. That's what Jeff Bezos did. He was saying some crazy stuff back when he was building Amazon in the 90s and he's changed his opinion. It will be very evident from new clips that, oh, that's an old one because it's very iconic. His older look now, his new look, he looks transformed. He looks very different.

[Tyler]

I don't think Jens has to go bald, but he can do the Bezos playbook, which is just get really jacked.

[Host 2]

Get really jacked.

[Tyler]

As he gets more jacked, you'll just see. Oh, yeah. I mean, he's looking like 320 lean right there.

[Host 2]

There you go.

[Tyler]

Like that must be pretty decent.

[Host 2]

320. Jensen at 320 would be electric. Well, he's.

[Host 1]

We have a video of the Air Force One Golden Palace. Okay, pull it up.

[Host 2]

What is the Golden Palace?

[Host 1]

That's the.

[Host 2]

Is that what they call the interior these days?

[Host 1]

The Air Force One that was gifted by Qatar.

[Host 2]

Okay. And that one is never going to be in use, but going to go straight to the library, I think. Is that the idea or is it going to get flown around? Well, let's. Let's watch the.

[Host 1]

Presumably it's going to be used.

[Host 2]

Golden Palace. Really particular aesthetics here. It looks like a normal 747 on the outside.

[Host 1]

Looks like a yacht.

[Host 2]

I feel like. Isn't there a big. There's a big swing in fuel efficiency based on the color of the plane. If you paint your plane black because you think it looks cool and it will attract more heat and you will burn a lot more fuel. And so most companies go with white. Spirit Airlines went yellow. I don't know if that had anything to do with their downfall, but you would think, is there a Chance. Is there some weird economic logic where if you paint it gold, the gold will reflect the sunlight and increase fuel efficiency?

[Host 1]

Unfortunately, the exterior is not going to be.

[Tyler]

Well, so you should wrap your plane in like a mirror, like.

[Host 2]

Yeah, maybe, but yeah, I don't know

[Tyler]

if that actually reflect the most.

[Host 2]

That'd probably reflect the most. Right. Mirrored plane. There are some silver. There are some silver bombers and jets that are in service across the United States Air Force. I can't call upon it right now.

[Host 1]

This really is a sky yacht chess board.

[Host 2]

You think anybody's played chess on there?

[Tyler]

4D chess?

[Host 1]

Yeah. The whole plane is actually used for 4D chess, John.

[Host 2]

4D chess. Chess is already 3D. By default, it's a 2D playing board. But time is the third dimension because the term that you're on is a dimension as well.

[Host 1]

This goes deeper than I thought.

[Host 2]

So 4D chess is merely a 3D representation of 2D chess that's played over time. That is the fourth dimension. Anyway, AI is being put to good use. Visualizing Seizure Ping, Donald Trump, Elon Musk and Tim Cook having a beer and I guess a cigarette as well in China somewhere. And then look who pulled up. If you scroll down, Jensen coming in from the back looking pretty good. Looking pretty good. Maybe not 320 lean, but he's certainly up there. He's scrolling up and people are really having fun with this. It turned into a little chain letter. Anyway, what let's go through the Wall Street Journal's report and what the editorial board had to say about the stakes of this summit. What's actually at stake? So the United States wants stability, but Chinese communist leader has larger ambitions, says the Wall Street Journal's editorial board. President Trump visits Chinese President Xi Jinping in Beijing on Thursday and the pre meeting US spin is a search for stability in quotes. It's a nice idea as long as Mr. Trump doesn't think personal rapport can overcome Mr. Xi Jinping's anti American purposes. Says the Wall Street Journal. The agenda ranges from trade to technology to Iran. On trade, the best outcome may be ratifying the status quo. A truce on tariffs with a promise from Beijing not to hold the world's rare earth supply hostage again. The markets, business leaders, no one wants chaos. Everyone can sort of work around a tariff plan if it's going to stay in for a long time.

[Host 1]

For Trump to be coming and being like, look guys, all I'm asking for is a little stability here.

[Host 2]

Yes.

[Host 1]

I just want stability.

[Host 2]

Yeah.

[Host 1]

I want calmness.

[Brandon Hill]

Yes.

[Host 1]

I don't want Any conflict?

[Host 2]

Yes. Didn't he famously call himself a stable genius? Stability has been part of his brand for a long time, at least the one that he projects. Mr. Xi Jinping, on the other hand, China's weak economy is an incentive for him to cooperate. And he'll Hope to placate Mr. Trump with promises of buying more farm goods. We talked about this previously. More aircraft and other things. So that's maybe why Boeing is there, although there is not. It doesn't look like there's representation from an agricultural leader, though I can't name one off the top of my head. But maybe they should be in, maybe they should be represented. But Mr. Xi Jinping has made that promise before. And US farmers have never regained their lost market share in China. Rare earths ransom offer. The rare earths ransom offer may be more US advanced chip exports to China. Xi Jinping views artificial intelligence as a decisive theater in Beijing's competition with the United States. And he is trailing, though not by much, maybe six months by most people's estimates. The administration wants to talk with Beijing about AI guardrails and by all means keep the phone lines open, but don't expect much from AI arms control. And the best deterrent is US dominance on models and computing power. Beijing will be happy to make pronouncements about responsible stewardship and then pursue its own interests with little regard for norms or laws. Beijing is engaging in, quote, industrial scale theft of American AI models. The Trump administration warned this year. And don't forget the Justice Department's indictment this year of a technology executive and associates allegedly running a sophisticated operation to divert high end chips to China. That was super Micro. Super Micro, yeah. They were doing the. The heater, what was it? The hairdryer.

[Tyler]

Yeah, they were putting the sticker shipping labels on there.

[Host 2]

Fake shipping labels. Yeah. Noticeably absent from this trip are the key leaders of the top AI labs. I mean, Elon Musk is there. Yes. And Jensen is there. Obviously deeply involved in AI, as are many of the other folks. Dean of Pal McCormick at METTA. But you don't have Demis from Google, you don't have Sundar from Google, DeepMind. You don't have Dario from Anthropic or Sam Altman from OpenAI. And each one of those leaders has their own complicated relationships with the federal government, the Trump administration. So it's not shocking that they aren't there, but it does feel like there's a little bit of a disconnect between the conversation that's happening in Silicon Valley and the conversation that's happening on the global geopolitical stage at the highest levels. And so if you were looking for answers to some of the biggest questions posed about how superintelligence will play out in the global stage, the AI 2027 scenario, the China wakes up scenario, you're going to have to keep waiting because that's not really what this, this, this is about. It's about maybe a little bit of semiconductor supply chain details and export restrictions, but not answering the big questions about what a US China relationship looks like in a post AGI world, a post ASI world, a post, you know, a fast takeoff scenario.

[Host 1]

OTP says. Why is Meta there at all? Good question. All Meta platform products have been banned for quite a long time. I don't think that will change. But they have this Manus thing. I think it might just be to show some friendliness, some presence there. Again, I'd be shocked if there's any time between Xi and Trump that gets dedicated to MANUS because like $2 billion is really a rounding error when you look at all of these other issues.

[Host 2]

But, but it is a big question for both economies. Not not only can American companies acquire Chinese companies that have relocated to Singapore and have sort of moved out. Like how, how strict will China be around maintaining talent and restricting exports of whole companies and technologies? They're even like a single sentence on it from this talk will clearly resound throughout the industry. And so the Venus fly trap that Xi Jinping is setting for Mr. Trump is on Taiwan. Xi Jinping wants veto power over arms sales from the United States to Taiwan, and he is pressing for the United States to formally oppose Taiwanese independence as opposed to the current posture of not so supporting it. Xi Jinping will argue the tweak is of no great consequence to Americans and Stoke and stroke Mr. Trump's ego that he can bring peace to one more troubled region. Yet that change would disrupt decades of US Policy that for all its delicate diplomatic wording, has held the peace. Taiwan is not the aggressor in the Taiwan Strait, a Xi Jinping fiction that opposing independence would indulge. Mr. Trump may not care about Taiwan's freedom or its example that a prosperous Chinese democracy is possible. But the president doesn't want a crisis on his watch, which would be an economic and geopolitical catastrophe that's 100% accurate.

[Host 1]

He would never want an economic geopolitical catastrophe on his watch.

[Host 2]

Xi Jinping will be looking to see if Mr. Trump suggests he won't defend Taiwan in the clutch. Trump's diplomacy is above all personal, and no one can predict what he'll do in the room. Japan and others in the region are watching with anxiety, a reminder that US Support for Taipei is an interest that informs America's alliances around the world. One mistake would be not stopping in Tokyo to advance Beijing in advance of Beijing as a signal of solidarity with Japan. Trump has said he'll bring up the case of political prisoner Jimmy Lai, but Xi Jinping won't move to release the 78 year old publisher who was convicted of bogus charges in Hong Kong, unless he believes Trump's request is more than a token gesture. The US wants China's help on Iran, and it would be an improvement if China at least stopped actively helping the enemy. Retired Navy Rear Admiral Mark Montgomery suggested last week that Mr. Trump should ask Xi Jinping directly if he's helping Iran with intelligence. And the Wall Street Journal goes on to say the larger context for this summit is that the CCP continues to be the main financier and industrial base for the world's bad actors from Russia to Iran and North Korea. The first Trump administration understood China as a strategic adversary, military, economic and ideological. The second Trump admin is searching for detent and Mr. Trump is the chief administration dove. This has some merit if America spends the interlude diversifying its rare earth supply chain and passing a $1.5 trillion defense budget to rearm, says the Wall Street Journal editorial board.

[Host 1]

Anyway, yeah, a lot of people are speculating that there's already been significant deals that have been made and that this whole thing is ceremonial because if you bring, if you make this huge trip and you bring all these leaders of industry and then you don't come away with some significant progress like it just feels like a even bigger l. So I'm expecting some, some real progress out of this. And, and it's good to, it's good to, it's good to see. I every time we talk to a China, to an expert on China like, like Bill or Jordan Schneider, it's always shocking like how little like direct comms there actually is between these groups. And so the more the better.

[Host 2]

Well, in somewhat related news, Anthropic has granted Mythos access to three major Japanese banks. Japan's three megabanks are set to gain access to Claude Mythos, the powerful artificial intelligence model developed by US Startup Anthropic. As soon as the the end of May, the banks Mufi Bank, Sumitomo Mitsui Banking Corp. And Mizuho bank were likely informed of the move by the US Treasury Secretary Scott Bessen in a meeting in Japan on Tuesday. This will mark the first time that a company from the East Asian nation has been granted access to Mythos, the AI model, which can discover and exploit software security laws far faster than earlier. Tech had been restricted to just 50 or so corporate corporates and organizations worldwide, including US firms, US banks, and UK government organizations. So this isn't the first international expansion since the UK did have access. But now Japan does. Japanese Prime Minister Sane Takechi had instructed cabinet members to ramp up efforts to find cybersecurity weaknesses in Japan's infrastructure and minimize risks posed by cutting edge AI models like Mythos. Access to Mythos was likely mentioned in her meeting with Besant on Tuesday. And so Chris McGuire says tonight, the Secretary of the treasury is personally vetting and approving each company that gets access to the most advanced US AI model because the risks of the model being misused to hurt US national security are so high. Also tonight, Jensen Huang is flying on Air Force One with President Trump to Beijing to sell China the AI chips it will use to develop its own Mythos level AI model as soon as possible. The administration. The administration's AI policy remains inconsistent and incoherent. It is impossible to justify these two approaches simultaneously. And, yeah, I'm still sympathetic to the export control take. At the same time, I feel like there's a larger negotiation that's going on, so I can sort of see both sides. But the interesting question that I've been noodling on that I haven't fully gotten to the bottom of is, you know, Dario has said that, you know, the open source community, China, is maybe six months behind Mythos, and that seems to fall in line with the progress that we're seeing from all the American labs. And so the interesting question is, what would the Mythos moment look like in China? Because in America, there was a showdown with the Department of War and Dario going back and forth and Emile Michael, and there was threat of a supply chain designation, which it doesn't feel like ever really went anywhere because the business is still cooking and it's available on all the hyperscaler clouds and the models being granted to Japan. And so it feels like the administration has backed off of that. But it was a really tense moment. And it did call into question all these debates around what rules and what authority should rest with the private sector versus the government. And America handled it in sort of a democratic way, loosely sort of evaluating the technology, having the option to sort of nationalize. And there's this bigger discussion around that. But in China, we've seen many times where tech companies have gotten really big and it feels like they, the CCP is applying pressure to the CEOs and high flyer is the company behind. Deep Seek is in a very unique position where the founder has an immense amount of control and you know, no

[Host 1]

matter what you think of 2 billion of his own money in a really low dilution round.

[Host 2]

Yeah. And I'm pretty sure he's a solo founder who like owns probably like majority control, at least majority voting control, if not majority economic control control. And that's wildly different than Anthropic where Dario had like 15 co founders, tons of dilution from, from different funding rounds. And you know, we had talked about the SPV thing yesterday. Like there are lots of people on the cap table. Like Anthropic is going the way of like a typical American company where there are lots of different stakeholders. They might be a public company soon, which means the government has more oversight as the SEC controls a lot of what they will do. And so you're in this odd scenario where you, you could have a similar level of technology deployed in China from a very closely held private company that is up against a much more aggressive government. And so what that looks like.

[Host 1]

Yeah, what that looks like functionally. Yeah.

[Host 2]

Like if you can't do E commerce without disappearing for a week or something, do you think you can do Mythos level AI cy, cybersecurity bug finding at the same time, while having full control and not being a public company, it feels like the stakes are much higher for when Deep Seek, if Deep Seek is able to catch up. There's some evidence that China's on a slightly shallower growth trajectory. But still, if you take Dario's claim of six months, even if you extend that to a year, like something is going to happen there where China is going to have their own Mythos moment. And what that looks like and what we learn about it I think is going to be dramatic and interesting. And also it's just, it's just odd that like this moment, there's this backdrop of all this stuff that's happening in AI in America and then there's going to be, you know, talks over like, you guys want to buy more apples? Basically.

[Host 1]

Yeah.

[Host 2]

Like, you know what I mean? Obviously there's going to be more high stakes to discussion around geopolitical conflicts and rare earths, but we are talking about like tariff rates at a time when just a couple months ago in America we were talking about like, you know, is AI going to like, be able to take over the government. And like, what does that mean? And how does the government control the AI technology as it diffuses? So very, very interesting things coming at some point. A lot of this will probably be pre negotiated behind the scenes since the Mythos moment happened in America and Xi Jinping obviously saw it and can get ahead of it. But you do see this consolidation of control at deep Seq and there's a question about, well, where does that go? Tyler, do you have any thoughts on this? Am I directionally correct on the timeline or.

[Tyler]

Yeah, I mean, it's concept seems like

[Host 2]

something's going to happen.

[Tyler]

Like there's the graph that shows that open source is increasingly getting further and further away from frontier models. But yeah, I don't know, it's just like so hard to tell, like, what's actually going on in China. Like what. What does it actually. Yeah, like what happens when the CCP starts to. To have, you know, more control over a company? Like, what does it actually look like? Yeah, in instantiation.

[Host 2]

Yeah. I mean, it's interesting. Like, so much of the way the US Government exercises control over the development of AI is on like the FDA regulatory. We want to review it, we want to take a look at it, we want to take it for a spin and throw some prompts in there before we let you send it to Japan, for example. And in China, it's a very. It's a much more industrial process where, you know, if the. If the chips are coming in and it's a whole negotiation at the government level, there's just the question of like, okay, can you even get the data center to. And it's much earlier stage, whereas that's not happening in America at all. Like, there hasn't been any movement on the government saying, okay, well, we're going to. The way we're going to control AI is that we will be. That we will decide if you can have a gigawatt. We will decide if you can have 10 gigawatts. Whereas in China, that's sort of the de facto status quo.

[Host 1]

Yeah, they. They still are very pro competition.

[Host 2]

Yeah.

[Host 1]

At least until you get these, you know, breakout winners. So, yeah, we'll see how that.

[Host 2]

And can they continue to be pro competition if they need to consolidate all of the Blackwells that are coming into the country into one data center center, this network together that can actually train a Mythos level model? Right. That's a big question. But people are having huge FOMO over the summit. Dude, I wish I was in Beijing with Elon Musk. Trump and Xi Jinping talking business, discussing AI Real boss moves. Must be a movie. Says someone who we're not going to read because it's a bad name in the thing anyway. Speaking of Elon Musk, SpaceX, Google are

[Host 1]

in talks to launch data centers in orbit deal between the two tech titans would give a boost to SpaceX business. Ahead of a historic public listing. Google is in talks for a rocket launch deal as the search giant expands its own efforts to put orbital data centers in space launch deal would put the two companies in partnership as they gear up to compete on orbital data centers, an unproven technology that SpaceX CEO Elon Musk has said is the next frontier for his rocket company. Google is also in discussions about a potential deal with other rocket launch companies. They got to at least have a stalking horse.

[Host 2]

How are we not talking about blimp data centers yet? I know Sergey Brin has a blimp company. Bezos has a rocket launch company. So there's a very logical line from AWS to space data centers through Blue Origin. But we got to put some GPUs on the blimps. We got to start mass manufacturing the blimps. This could be the thing. I mean, there's no nimbyism in the middle of the Pacific Ocean if there's just a bunch of blimp data centers floating around in a circle.

[Tyler]

Well, I think if you're already in the ocean, you should go down as well. You know, Atlantis map cooling. Yeah, I feel like none of the ocean guys that are working on ocean robotics are thinking about like putting ocean at the bottom of the ocean.

[Host 2]

Yeah, I mean, it feels easier than space data center insurance. The space is free, the land is free. You don't run into the regulatory stuff. Cooling, potentially free energy. Little bit tricky down there.

[Tyler]

Geothermal.

[Host 2]

Geothermal, maybe. I don't know.

[Host 1]

No, there's oil and natural gas reserves

[Host 2]

on the oxygen there. Cause how do you burn the oil in the natural gas water? Big, big diving bell to be a scuba. Scuba tanks or something. Anyway, let's continue. Where are we in this? SpaceX and Google.

[Host 1]

Speculative technology has been at the center of SpaceX's pitch to investors ahead of its planned IPO this summer, which is anticipated to be the largest IPO of all time. Last year, Google announced its own plans to launch Prototype satellites by 2027 as part of a moonshot initiative called Project Suncatcher. It's working with another company, Planet Labs, to build those satellites. We're working on getting Will Marshall from Planet Labs on the show in the next week or so, we'll send tiny racks of machines and have them in satellites, test them out and then start scaling from there, says Sundar. There's no doubt to me that a decade or so away we'll be viewing it as a more normal way to build data centers. A little bit of an interesting timeline. There's no doubt to me a decade or so away we'll be viewing it as a more normal way. Yeah, a decade or so away.

[Host 2]

He's being a little cautious in his language on his timelines, but Google was

[Host 1]

an early investor in SpaceX and owns printing.1% of the company. Not bad. The other funny dynamic from SpaceX is dealmaking recently is given that the OpenAI startup fund was an early investor in I believe Cursor's seed round. True $8 million round OpenAI. At least the startup fund will have some meaningful exposure to SpaceX.

[Host 2]

Yeah, SpaceX, an OpenAI backed company. That's the way people will introduce it. OpenAI backed SpaceX IPO?

[Host 1]

Yeah, I thought it was, it was quite interesting having that happen while the whole trial was going on.

[Host 2]

Everyone owns everything, everyone has diversified into every, every company at this point, basically. Let's see. The company last week announced a deal to sell Earthbound computing resources to the AI company Anthropic. As part of the agreement, Anthropic expressed interest in working with SpaceX on orbital data centers. Earlier this year, SpaceX filed an application with federal regulator to launch up to a million satellites for its orbital data center ambitions. That's crazy. I remember, didn't they get what was the original Starlink approval? I thought it was like tens of thousands, maybe 100,000 satellites. I'm starting to think like looking ahead, there's going to be a backlash to their blocking out the satellite. Like we've seen the Matrix. This happened in the Matrix, but for like chemical reasons. But like if there's like light pollution from this at all, it's going to be a absolutely massive. Oh yeah, Panthalassa Ocean data centers. You don't need to sink them to the bottom of the Mariana trail. How can we forget do them on the, on the ocean. Did you look up how many satellites SpaceX currently has in orbit? Because I think, I mean it's a huge number. It's over 10,000 I believe. But a million is a huge jump. But that should be, that actually should math out. I think we did the math. That should math out to gigawatts in space, which is definitely meaningful.

[Tyler]

So right now they have like just over 10,000 satellites.

[Host 2]

Just over 10,000. Okay. But I think they're approved for maybe up to 100,000 or something. And so they go and get the huge, the huge allocation and then they start launching. Very, very excited to see the progress on Starship. Isn't there a launch that happened or is happening this week?

[Host 1]

Yeah. When's the next time? I'll keep reading the other story. This was back in March. Blue Origin has a play in space data center as well. So I would kind of. You could imagine them and Amazon teaming up in some capacity. But you could also imagine Amazon saying, like, hey, if we're going to be a real player here, we need to, we'll probably need to work with Moon.

[Host 2]

Star Cloud finally probably finds a dance partner, if not multiple. Many industry leaders see orbital computing as a solution to the limitations of earthbound data centers. Earthbound is capitalized, it's a proper term now, which require swaths of land and power. These data centers are designed to be powered by solar panels emitting, eliminating the major power constraint faced by data centers on Earth and one of the major ecological issues with emerging technologies. But there are also major engineering challenges. SpaceX made its name as the world's leading private launch provider, sending NASA astronauts to the space station and launching thousands of satellites as part of its Starlink Internet constellation. Over 10,000, apparently. As it prepares to go public, SpaceX has struck a number of deals that have rewritten its balance sheet, helped Musk consolidate his companies, including its acquisition of Xai. And SpaceX also announced a deal with Cursor, as we have talked about on the show

[Delian Aspro]

Starship.

[Host 2]

What's the update?

[Tyler]

So the next launch, the 12th one is next Tuesday, the 19th.

[Host 2]

Next Tuesday.

[Tyler]

That's when they're targeting.

[Host 2]

Okay, yeah, yeah.

[Host 1]

And what's the goal?

[Host 2]

But they were like moving it down the Runway, sort of getting ready to stack.

[Host 1]

Do you have any idea what success looks like?

[Host 2]

I think this is the first orbital

[Tyler]

test of the V3.

[Host 2]

Okay, so they got to get it to orbit and then get it back and probably land it. But are they going to land on land or water? Because didn't the last one make it up and back but exploded in the water? Or was that the Blue Origin New Glenn? I don't know. Anyway, there's a whole deep dive on space data centers. In the Wall Street Journal. They say, is it a pipe dream or AI's next big thing? They're pointing out some of the challenges. You're going to need thermal louvers opening and Closing shutters to block or allow heat thermal contact. Special bolts to separate items with different heat requirements. Deployable radiators. Heat is transferred to special surfaces that unfold in orbit. It looks like a big pull up.

[Host 1]

This graphic showing sun synchronous orbit.

[Host 2]

Yeah this is cool. So when, if you're orbiting the Earth vertically, essentially the sun is constantly hitting the satellite. So you're not out of power for half the day. But they say the radiators for data centers will likely be very large. I like that. Phase change materials. Materials that can absorb. They can change from phase to absorb heat like wax or salicylic acid. Usually from solid to liquid when returned to normal when temperature is reduced. Sun Synchronous Orbit 373 miles to 497 miles up. So about 400 miles up is when you get into sun synchronous orbit. That's on the low end of low Earth orbit. Low Earth orbit around 1,000 miles. Mid orbit 1 to 10,000 miles. High orbit over 22,000 miles. And so we covered a decent amount of this so we can move on. Do you want to talk about Shanu?

[Host 1]

Matthew was adding the conversation we had with Doomberg.

[Host 2]

Doomberg was fine.

[Host 1]

Doomberg was saying the only way the data center build out will be successful is by either breaking all the rules or going off grid. And Shanu said the tech crowd keeps leaning into this off grid is the only way thesis. And I disagree with this notion. The exhibit example of XAI Colossus even overstates what's actually happening at Colossus. XAI did not build a permanently off grid AI campus because the grid was unusable. Colossus already has 300 megawatts of TVA interconnect approved across two substations. XAI funded plus contractual obligations to curtail during periods of grid stress. That is a grid tied architecture with supplemental on site generation, not an islanded power system.

[Host 2]

I think, I think Doomberg was like the other side of this. That Doomberg did highlight was like Colossus is the exception that proves the rule because it moved so quickly and, and took advantage of so many different grids. And there were power generation from multiple states at one point and now there is pushback. Like there wasn't pushback when Colossus was built originally because data centers were not a hot topic. And it's the same thing with launching rockets. Like we were talking to someone who lives up in Santa Barbara. They're launching a lot of rockets from Vandenberg. It's getting to be a lot of traffic. It hasn't risen to the level of protest. But you can imagine that building 10 more Vandenbergs would be difficult because you go to the local community and you say hey, we're going to build a space launch station, a space port. And they say oh, that sounds good. Like you mean because we're going to go to the moon once a decade? And you're like no, we're going to launch rockets 10 times a day or 100 times a day. And people if they have an example that they can draw from and so say oh well, like that community doesn't like it, so I shouldn't like it. That's some of the pushback. But anyway, we can keep reading Shanu

[Host 1]

Matthew's post yeah, let's see.

[Host 2]

Turbines solved a time to power problem, so it accelerated the deployment while substations and transmission infrastructure were being built. That's very different from do all of this off grid and economically most large scale AI load will still want to be grid tied load long term. Interesting. I have made this point several times, but running your own generation fleet at Hyperscale is expensive and operationally complex. Fuel logistics, maintenance, N plus 1 redundancy permitting, emissions compliance, balancing load and supply continuously, multi year equipment lead times. You don't just figure that out or want to do that overnight. Amin the chief Technology officer at or the chief technologist at Google AI. Does that mean DeepMind straight up confirmed this on a podcast and they literally bought a hybrid island power developer. BTM generation absolutely matters. It's becoming a critical bridge solution during a period where utility interconnect timelines are badly lagging demand growth. But I'd strongly push back on the idea that the steady state future is fully off grid hyperscale campuses. The industry is converging towards hybrid architectures, grid interconnect plus captive generation plus storage plus demand response. They're throwing the whole kitchen sink at it. Anyway.

[Host 1]

Related yes, more Perfect Union had a post earlier and they're turning off.

[Host 2]

They're turning off.

[Host 1]

Nearly 50,000 people in the Lake Tahoe area have been told their utility will stop providing power to them because it's redirecting that power to data centers. NV Energy, the Nevada utility that supplied most of Lake Tahoe's electricity for decades, says that next year will stop servicing homes in the area and instead direct that electricity to the growing demand from Nevada data centers. Northern Nevada is one of the fastest growing data center corridors in the country. I read this, I was like this is completely insane.

[Host 2]

It's so stupid. Like, you read it and you're like, this is such an L in the making for the data center industry. What were they thinking?

[Host 1]

And then it's immediately getting community noted in a quote from the president of Liberty Utilities in Lake Tahoe says this does not mean the power is shutting off. Yeah, energy companies, utilities and large customers change energy supply frequently. So there's a supplier of energy that is saying like, hey, we're not going to renew our contract. We're going to divert the energy elsewhere. But there's a bunch of suppliers of energy. And so Andy Masley says everything MPU posts about data centers is complete garbage. They have zero respect for their audience. Literally no one here is losing power. This tweet is a complete lie. What's actually happening is that a supply contract between two utilities is ending and the small one is just buying power from elsewhere. And this was all expected to happen since 2009.

[Host 2]

No way. Wow.

[Host 1]

The company that serves, how you're going

[Host 2]

to say, like two years ago, the

[Host 1]

company that serves homes on the California side of Lake Tahoe is a small utility called Liberty. Liberty buys about 75% of its electricity from a much larger utility, NV Energy in Nevada, and generates the other 25% itself from solar farms it owns. Liberty sells, then sells that 49,000 to 49,000 customers. NV Energy has told Liberty it will stop selling them wholesale power after May 2027. It's kind of like Liberty's economy coffee shop that buys beans and sells coffee to customers. The customers are the homes and the beans are the electricity it buys from NV storage and or makes itself. This is like your local coffee shop ending a contract with a specific bean company and started buying beans from somewhere else. It doesn't stop you from buying coffee. Why is their contract with Envy? Why is their contract ending with Envy Energy? NV Energy selling to Liberty was understood as transitional since it started in 2009. Long story short, NV Energy was basically Liberty's only wholesale option. But a new transmission line opening in May 2027 gives Liberty access to a much wider western market and among other things, a much larger share of solar and wind and hydro. That's the whole story here. Ending the contract with NV Energy and opening up this much wider pool with much more renewable energy. Was the plan here completely separate from data center demand?

[Host 2]

Interesting. Well, anyways, it would be remarkable if all of Tahoe lost energy and Mark Zuckerberg couldn't charge his electric wakeboard because

[Host 1]

I somehow he would hit the streets he'd be protesting.

[Host 2]

I think he'd be protesting. I think he would get something done if power was lost to Tahoe. I mean, a lot of influential people are out there.

[Host 1]

So anyway, all right, we got to talk about on that. We got to talk about unitree.

[Host 2]

Unitree.gd01 they build a new robot and you're going to love it. It's not scary at all. And it looks very fun and safe. This is from Avatar. They just built the mech from Avatar. This has been pictured many times in science fiction and will strike fear into many people.

[Host 1]

Coincidence like that they use a racing bucket.

[Host 2]

So it is cool. It does feel very hacker culture. Like it feels like this was like a crack team, a little skunk works project. This is. It doesn't look like as polished or ready for, you know, sale, but remarkable results.

[Host 1]

They're like, you got to put this out so that the Americans cancel all of their technology ambition and just go back to the Dark age.

[Host 2]

It is intimidating. Is it a coincidence that it was launched the same day that the same week that the summit is happening in China? I don't know. It is amazing that it turns into a horse of some sort. A four legged creature. Although how do you drive it at that point? If you're laying backwards, it needs some sort of gyroscopic seat. That one hasn't really been considered because if you're laying back while you're driving it in horse mode, you. You're going to be. Maybe you'll just nap. Maybe you're like, oh, I want to commute. Henry Ford said if I asked people what they wanted, they would ask for a better horse. This looks like a better horse to me. You get to work, you pop up into vertical mode. I'm fascinated by how they're controlling it. Is it just like a video game controller? Because some of this seems like it would require you to telegraph in. In the Avatar films, the operator of the mech has full. The input is the actual arms. So if they want to swing a sword, they swing the sword or they swing the machete physically with their arm. And then the robot does the bigger

[Host 1]

version of that timeline. Until these mechas replace horses for police use cases.

[Nate Tepper]

I don't know.

[Host 2]

There's gonna be pretty crazy pushback if that's rolling around during a riot.

[Host 1]

I know, but the benefit here is that I don't think the Mecca will go to the bathroom on the ground.

[Host 2]

Obviously this is gonna happen at some point. A lot of these developments are very early. I bet you if you give that thing a strong kick, it falls over and sort of gets stuck.

[Host 1]

At this point, Sayla over on X had pulled out the funniest part, which is that Unitree unveils the world's first mecha. Show it breaking down a brick wall. And then in the launch announcement, they say, please everyone, be sure to use the robot in a friendly and safe manner.

[Host 2]

Yeah, Wild. Egocentric data collector versus the Mecha mounted teleop. Chad. Absolutely wild. Wild. Is there another video of this? Oh, yes, this is an AI video, but we have to play the version of this getting hot. Dropped in into what you could imagine as the battlefield. With four FC 204 drone formation, max payload of 600kg, Unitree's new Transformer mecha, 500kg of weight. That's probably without a human inside. So you might need six drones. But this is something that feels not that far away. Not that far away. Although you drop that thing in the ocean. I feel like there are so many steps on the humanoid robot path where the deployment in a flat surface in a warehouse makes so much more sense than standing on a ladder straddling a chimney in the rain. And there's a little bit of mud and you use a power tool and dust kicks up. Actually, ruggedizing all of this is probably a whole second step in this process where you're adding more weight. That means more battery. You need better battery technology. I don't think you'll be dropping this into the Taiwan Strait anytime soon, but it certainly strikes fear into the hearts of many.

[Host 1]

Well, there's more humanoid news. Brett Adcock has a live stream going.

[Host 2]

That's right.

[Host 1]

Of their robot.

[Host 2]

Yeah.

[Host 1]

I do wonder how they're. Again, I'm sure the comments are all, how do we know this is actually live?

[Roman Chernin]

Right?

[Host 1]

Like, how do you. How do we know?

[Host 2]

So is this the actual live stream right now? It's continuing going because I pulled this up an hour ago and it looked, I mean, I guess exactly the same because it's the robot and they've been sorting for an hour battery pack or plugged into the wall. What do we think? What's the battery life?

[Host 1]

He says, watch a team of humanoid robots running a full eight hour shift.

[Host 2]

Okay. Okay, so they might have swapped it out at some point for a default.

[Host 1]

Yeah, they could just trade.

[Host 2]

Interesting. It looks like a lot of progress.

[Tyler]

It's extremely fast.

[Host 2]

It is fast. Like the movement's very natural. I wonder what is the goal here exactly? Because is the robot sorting these. What does it do with the blue one?

[Tyler]

I Think it's putting it so the

[Host 2]

shipping label is visible or the shipping labels down. Okay, okay, okay. Yeah, that makes sense. That feels like. Because some of the robot applications are things that you could do with a two axis gantry or a sorter. Like if you watch how it's made, there are industrial machines, not robots that can do all sorts of sorting. Like you watch like how Diet Coke gets filled and. And the cans roll down and at some point they all need to be facing upwards. And so there's a machine that sorts them and tilts them all upwards. Obviously the humanoid use case is much more unstructured and that is beneficial for so many things. I don't know if this is.

[Host 1]

Apparently they're just going to have the robot go until it taps out.

[Host 2]

Ooh, okay, I like that. That's great. Very cool. Well, in other video news that we need to play, General Catalyst launched an ad on X that might be making it into TV ads or TV ads at some point. What?

[Host 1]

Gcat.

[Host 2]

Gcat? What's gcat? Oh, Catalyst. Maybe that's the new ad campaign. But the campaign is called Meet gc and it's a clear reference.

[Delian Aspro]

I'm gc.

[Host 2]

And I'm vc.

[Delian Aspro]

Who's your friend here?

[Host 2]

Vc, this is Woof AI, an AI native companion platform that combines robotics and machine learning. You'll never want a real dog after this.

[Delian Aspro]

Well, I think people like dogs as they already are, though.

[Host 2]

Vc, you don't need to walk it. You never need to tell the kids you sent Milo to the farm. We're leading the seed and could probably make room for you.

[Delian Aspro]

I'd love to hear more, but we actually have a really high bar around responsibility for the these things. Is Wolf AI okay?

[Host 2]

Of course. He's fine. Oh, sorry, buddy doesn't want. Easy, easy. Stay, stay. No, no, Stop, stop, stop, stop. I'm sure he's interesting symmetry to the podcast set that Reggie's doing, right?

[Host 1]

Background Reggie directed.

[Host 2]

Yeah, but it's cool that like the brand, like you could wind up doing two very different designs. But there's like a. There's a through line there. Even though it's obviously a nod to the Mac versus PC ads from the 2000s, which we can pull up to watch one of those. I think we've watched these before and they have a very similar stiff PC versus Mac look. But I think that was a good message of like, okay, let's not break the other cat. But also let's not fund the stuff that is pointless and useless. And if you already have A dog and keep a dog.

[Host 1]

This is a big moment because you have been wanting a VC firm to just commit to paid advertising for a long time.

[Host 2]

Well, I wouldn't say this is a commitment yet. We gotta actually see spend. Yeah, of course, of course. Yeah.

[Nate Tepper]

No, no.

[Host 1]

Magazine ads.

[Host 2]

Yeah, in the right. I mean, certainly in like the Wall Street Journal, in Forbes and like the places that people, the entrepreneurs go for information. It makes so much sense to actually spend on advertising to build a brand. Especially when you're at the scale of General Catalyst. Can we pull up the Mac versus PC ads? We're not gonna watch all of them. There's 66 of them. Talk about a generational run a lot

[Max Levchin]

of the same kinds of programs.

[Tyler]

Yeah.

[Host 2]

Like Microsoft Office. But we retain a lot of what makes us us.

[Max Levchin]

You should see what this guy can do with a spreadsheet.

[Host 2]

It's insane. Life stuff like music, pictures, movies, stuff like that.

[Delian Aspro]

Whoa, whoa.

[Host 2]

What, what exactly.

[Max Levchin]

What exactly do you mean by better by better?

[Host 2]

I mean making a website or photo

[Max Levchin]

book is easy for me, and for you it's not.

[Richard Socher]

Oh, oh, that kind of better.

[Host 2]

Yeah, I. I was thinking of the other kind.

[Host 1]

What other kind?

[Host 2]

Hello. We can switch away because you've seen all these. Marc Andreessen quote tweeted it. Dunked. He said, stay tuned for our upcoming ad campaign. We're the VC who doesn't sneer at your idea. Interesting. Getting spicy. Yeah. Red alert over there. Red alert. It's a lot of red alert going on. It's a long thing.

[Host 1]

Well, both companies are invested in robotics company Anduril.

[Host 2]

Oh, yeah. I mean, I'm sure there's a ton of portfolio overlap.

[Host 1]

Oh, no. I just think like in another world that that's an Anduril product. That's dog thing. I'm sure they've experimented with that form factor.

[Host 2]

They've got some fresh funds to build a Mecca. It is. I mean, if there's ever someone to build the Mecca, it would be Palmer Luckey. He has talked about this stuff. I even asked him on stage once, like, walk me through the Pacific Rim vision, like, how do we build that? And we had a fun back and forth about it.

[Host 1]

Well, do you remember, do you remember much about his response?

[Host 2]

Yes. So his response to stuff like that,

[Host 1]

because the Mecha form factor is just like, hey, I'm a sitting duck on a platform.

[Host 2]

Yeah, exactly. So his point is always like, the technology to build the Mecca is way behind the technology to remote control something or put an AI on board or something like that. So you actually want to build the Iron man suit because you can just build the drone, the submarine, the automobile.

[Host 1]

Or he'll use the example of like the Batmobile. It's like, if you need a vehicle for the city, like, do you really need this massive vehicle? Right.

[Host 2]

Clearly Bruce Wayne, the parking, he's obsessed with building Batmobiles. It's very funny. I don't know. That is a very cool demonstration though. But yeah, Andrew always says don't build the Batmobile and focus on like, is there an actual, like urgent need from the military? Is there support from Congress to actually appropriate the funds and unitree? I don't actually know. That feels like just like an, that's like a marketing project at this point.

[Tyler]

Yeah, I mean, I think their, their IPO is like very soon. Right. Because I think the original application was like late March.

[Host 2]

That would probably pump it. And then also they're showing it like break down a brick wall. But when I was thinking about what that's actually valuable for, we were talking about could you deploy that and sneak through a forest and then attack the enemy or something? It's so big, it's so visible. A single drone with a fiber cable coming out of the back is going to blow that up. It doesn't really seem that practical. Maybe if you're going over really rough terrain, you want the legs instead of a tank tread, but you can sort of just put a huge tank tread on something and just crush everything in your way. But if you're building and you need to get a bunch of two by fours onto the roof, like there's a theory that that might be a little bit better. You were saying you're forklift maxing and you think the forklift will always defeat that. But there's got to be certain areas where you can't necessarily bring a forklift because there's a narrow, it's an uneven ground, a narrow entryway into the back of a house. You have all the two by fours you need to get them on the roof. You would otherwise be going up and down a ladder. And so Mecca to the rescue.

[Tyler]

Yeah, I mean, it's just like pretty hard to justify this, I think. Yes, for certain emergency scenarios, it makes sense to have some super weird emergency.

[Host 2]

I need a really cool thing, piece of tech. I need something that looks really cool before my IPO emergency. Well, we can, we can flip over to the other financial news that's making headlines on the COVID of the Wall street journal. Inflation accelerated 3.8% last month. Not good news. Joe Wiesenthal has a couple posts, but one is just man and he says the headline US April producer prices rose 6% year over year. The estimate was 4.8%. So obviously the closure of the Strait of Hormuz is having an effect. We were talking about Diet Cokes. You would think that the caffeine doesn't come from the Middle East. The delicious caramel flavor, whatever else they put in this does not come from the Middle East. But aluminum. Actually there are smelters in the Middle East. Aluminum is in shorter supply because of the closure of the Strait of Hormuz. And so Diet Coke is rapidly going out of stock. And so you need to load up if you got prepared.

[Host 1]

How many you got? Four or five pallets?

[Host 2]

I have not actually gotten any. I'm not stockpiled Diet Coke yet. But who knows, maybe prices go up. That's what inflation represents and Wall street is getting more anxious about it. So we can read through this quickly before we are joined by Max Levchin from Affirm. So surging energy energy prices have pushed investors inflation expectations to multi year highs. Tuesday's hot readout on consumer prices is intensifying Wall Street's anxiety about inflation. Even before the latest release of the Consumer Price index, inflation expectations had been climbing, a potential trouble spot in the market where stocks have largely shrugged off the US Iran conflict and the resulting energy shock. And so investors bet on inflation by buying and selling both ordinary US Treasuries and those that hedge the risk of inflation, called tips. Treasury inflation protected securities. The gaps between the yields of those two types of bonds, known as the break even rate, recently hit the highest level since October of of 2022, according to federal Reserve data, suggesting that investors expect annual inflation to average around 2.7% over the next five years. Now that's not entirely doom and gloom, but the risk here is that the AI boom, the GDP numbers are sort of concentrated in AI capex and data centers. It's not evenly distributed. The real economy is only growing at 0.1% while the real economy is suffering from 2.7% inflation over the next five years. That's a recipe for stagflation, stagnation plus inflation, a very rough thing for any economy to go through and something that the market and the economy and the politicians should be focused on. So the 10 year break even rate has also climbed hitting 2.5% this month, its highest level since 2023, driven largely by the war fueled surge in oil prices. The rise in inflation expectation worries investors because of the challenges it poses for the Fed interest rates policy and Major index's latest run to records. And so for a while, everyone was thinking about maybe interest rates will come down, maybe mortgages will become more affordable. But there was even the whole debate or the fight with Powell and lawsuits, which were eventually dropped. But Powell and Trump were at each other's necks and issuing statements about their vision for the future of the American economy. And on the flip side, you have Kevin Warsh coming in, maybe more friendly, maybe more receptive to a rate cut, but the data makes that very, very hard to justify. And so we are in.

[Host 1]

Yeah, in some ways, you know, who knows how history will view Powell, but like, he landed. He landed the plane. And then, you know, we get war in the Middle East. And now Kevin Warsh, in some ways, looks like he'll have just as tough of a job.

[Host 2]

Potentially. Potentially. Well, our next guest, Max Levchin from a firm is the co founder and stuff CEO. He's returning to the show. We've been keeping him waiting far too long. Max, how are you doing?

[Max Levchin]

I'm great. Thanks for having me again. Third time's a charm.

[Host 2]

Third time.

[Host 1]

What have you learned about third time?

[Host 2]

Can't be the charm. I want 25th time to be learned. Keep it going. I love these.

[Max Levchin]

I enjoy this course.

[Host 1]

What have you learned about making coffee since the last.

[Host 2]

Oh, yes, the important update, because I'm

[Host 1]

assuming you're never like, oh, I'm. I'm pretty good at this.

[Host 2]

Or are you resting on your laurels? Are you, are you the final boss?

[Max Levchin]

I'm absolutely not.

[Host 2]

Okay.

[Max Levchin]

I took apart E61 group, which is the classic 1961 sort of that. That's the granddaddy of all the espresso. Espresso group heads. And I wanted to learn how the mushroom valve works because I thought mine was clogged and it turned out not to be clogged. But I went down an extremely deep rabbit hole of taking apart an E61, which I've actually never done before. So I'm now really boned up on how espresso machines worked in 1961 and since.

[Host 2]

Yeah, yeah. People always talk about, like, when you get into coffee, you go. You effectively vertically integrate or go deeper in the supply chain. You start roasting the own beans. You gotta grow your own coffee. At a certain point, you have to be smelting the metals that go into the machines, understanding the alloys, coming up with new chemical processes. You have to set up your own mind.

[Max Levchin]

You're putting all kinds of wrong ideas into my Mind.

[Jubin Murzadogan]

Ooh, not yet.

[Host 2]

Yeah, yeah, yeah, yeah, yeah. Did, did you smelt the metal that went into the espresso machine?

[Max Levchin]

I've not smelt anything lately.

[Host 2]

I'm a novice, a real novice.

[Host 1]

Last time.

[Host 2]

Anyway, we're not here. Next time I'll tell you all about my smelting experiences. Fantastic. I imagine that the smelting process is as intricate and as rewarding as the coffee making process in some ways.

[Host 1]

Yeah. I expect you to be able to try a coffee and understand the alloy of the machine.

[Max Levchin]

Yes, possibly, but I'm very, very long depth. Anything that you can go very deep into, like this is the time in general. But like depth of the competitive advantage is like a profound strength. So the reason I'm so into whatever things I'm into is I found over the years that if you out depth your competitors, they just can't beat you. So I'm very pro smelting. I'm very pro going over.

[Host 2]

Yeah, yeah. That's interesting. Is there, is there a translation or if you transpose that to sort of like advice for young people who might have anxiety about traditional career paths? I was sort of, you know, we look at like job numbers all the time and layoffs. And at the same time I think about when I was a kid, if I said, oh, when you grow up, you will be a live streamer who. At an AI lab. It's like none of those work words existed back in the 90s. And here I am also a lot

[Host 1]

of, a lot of. There's so much advice out there. It's like you want to be a generalist. This is the age of being strong.

[Host 2]

Maybe it's the opposite.

[Host 1]

And when I look at people that are maybe sort of midway through their career the highest earning, the most respected are almost always the ones with extreme depth where they can simply out compete anyone else in that role because they know more about it, they put in more hours.

[Host 2]

Yeah, but, yeah, but what are you thinking for young people?

[Max Levchin]

I'm definitely big on depth. I think these are entirely non contrarian opinions, but maybe they're contrarian right now. But I think it's a great time to get a computer science degree. I think if you're ultra deep into really understanding how software is made, you are like, everyone's gonna be a 10x engineer. If you're 1x engineer yesterday, you better be a 10x engineer tomorrow. That's the new baseline. But if you really, really get it, if you're like, if you're smelting your own code, I'm just gonna go with the smelting. But if you're that good, you're gonna be 10,000x engineer and you will be worth your weight in smelted gold. Who knows? But it's very, very powerful to be a deep expert. Like you are the one AIs want to learn from. And like that, that is, that is the unattainable Mount Olympus of value. And so I, I would strongly suggest that being deep is great. Majoring computer science is great. If you love computer science, like this is the time to major in whatever it is you see yourself being as deep as possible because then you will become absolutely. What about valuable worlds?

[Host 1]

What about industry specific depth? Like if a founder comes to you and they're like, oh, I've been building, I've been building an E commerce, but really. And I've done quite well, but really I'm interested in space, I want to work on that. You've done, you're interested in space and

[Max Levchin]

you're not working on it. Like what are you doing? The very best time in history to do that.

[Host 1]

But generally you've done quite well by just focusing on something in a category that you know better than. Maybe there's maybe like three, four other people in the world that understand the intersection of like money and technology the way that you do. And I feel like a lot of founders oftentimes will do one thing, feel like they kind of like climb the mountain. And then some will just go back and climb a very similar mountain and do quite well again. Others want to jump ship and do something completely different.

[Max Levchin]

So I tried both. My last startup in between, the other payment startup was not in payments and it turns out that I'm very good at climbing this one mountain. I'm just going to keep climbing it because the reason I'm good at it is because I love depth and I love getting deeper, deep, deeper and deeper into payments, into how to make payments more expedient, more transparent. In the case of a firm, payments are a lot of things and access to credit is a totally new idea. Like I've never touched credit until a firm and now maybe without false humility, I think we are definitely one of the very best credit underwriters in the world and are just getting better and better. We just announced yesterday we built an entire new model family loosely inspired by the attention mechanism that is powering all of your alums out there, including your parent company. Now, some ideas we borrowed from all these really amazing discoveries and sort of pure language AI. And yet it feels like we're just getting started. Like every time I look at the sort of corpus of things that we can try to do, things we can build, the products we can launch, like, oh my God, we have so much to do. Like, there's absolutely no chance I'll get to rest. I'm not sure I'll get to. After all, it's the greatest time to build stuff because everything is faster and more exciting and easier to start. But the depth reveals itself the deeper you dig.

[Host 2]

Okay, so yeah, we sort of mentioned, we referred to it a little bit, the 2026 Investor Forum. Who attended, what was the goal of that? Take us through that event and then I have a whole bunch of follow up questions about how you're communicating at the current time.

[Max Levchin]

Yeah, it was amazing. I'm still kind of slightly high on the whole thing, but the thing that was really fascinating. So we did this event three years ago or similar event three years ago and it's like a blink of an eye. Like I remember it was in the same building, a lot of the same people attended. So these are mostly buy and sell side analysts, a bunch of our shareholders, a couple of board members actually flew in for which was like a nice surprise to me. And the entire management team gets on stage and like, here's what I work on, here's what the company is and but like the reason analysts are there, they're like, okay, tell us about your vision Max. Yada, yada, yada. Really, what will the next three years worth of gross targets going to look like? And so the main act, I was the warm up act, I tried to be funny, but the main act was our CFO Rob who got on stage and said, all right, so we're going to grow to $100 billion of GMV. We're just in the range of touching 50. We're going to double and we're going to do this by compounding at 25% every year. And we're going to move our profitability target from 3 to 4% which is our current range to 3.75 up. And so we're going to up the floor. We're going to be more profitable and grow faster. And it's like faster. We're a bigger company now. Like what are we doing? Like wait a second, last time we did this three years ago we said we're going to compound at 20% but instead we did it like 30 plus. And so just like this out of body experience of like we're getting bigger, the room was significantly larger, same building different floor, different room, many more people, many more shareholders. And yet we're telling people like, hey, we, we're going to compound that much faster than the last time. It's like, I guess it's like a flywheel. This network we've built. It keeps on spinning and spinning faster. And so it was amazing in a sense that I knew all these numbers. I obviously improved and worked with the team to make sure we feel great about them. But when you sort of say that out loud, compare that to the last time we told you the same thing. We're going to accelerate. It's like, wow, that is so cool. So anyway, so I'm still very high from being able to say that out loud is very fun.

[Host 2]

So there's a flywheel. I imagine with a lot of financial companies there's economies of scale and so that seems doable. Are you also talking about TAM at this point because the company is so large or is it still early enough that that's not something that investors are coming to you asking to sort of justify? Well, you know, the global economy is only this big and if you take 50% of it, well, there's nothing else.

[Max Levchin]

The good News is that $100 billion of payments in this country is but a drop in the bucket of an even larger bucket. We're in no danger. The US revolving credit, which is obviously a fraction of total economic rotation is like a trillion three. Right now, 100 billion is like, then we're taking shares. So I'd like to compare to that number. People are not using revolving credit, they're using a firm. Good for them. Non revolving, no fees, all that. And so we're not yet at the sort of town. By the way, these are all US numbers and we're already live in the UK and Canada and we've announced a bunch of other countries.

[Host 2]

Okay, so speaking of the investors, obviously it's very important to communicate with investors right now, both from your growth plans and also what's going on in the market. As you think back to the March time period, the SaaS apocalypse, every company got sort of beaten up. You came back really quickly. How much of that was show versus tell? How did you think about communicating to investors through that? How important was the investor communication versus just continuing to deliver on the actual metrics as you're confronted with one of these, like, oh, there's a new narrative. What's your playbook for actually working with investors to get them comfortable?

[Jubin Murzadogan]

You know,

[Max Levchin]

to be completely honest, I don't know. I think, well, I'm Usually honest, so I might as well tell the truth every time. So this particular time we said nothing. We said, okay, I guess the sky's falling. Doesn't feel like it's falling on us. We're printing a really, really good quarter. Yeah, we're about to report it. We could like wave our hands. You got it all wrong. We're fine. Or just wait a few more weeks and like, hey, here's what we printed. How about that? And by the way, we're going to guide so we're going to reveal that we're having a pretty good quarter too. And so that's what we did this time. And we just didn't spend too much time communicating how the story isn't true. Yeah, I don't know if that's like the best response. Like we've definitely in the past we would read, you know, the, the most recent time when I thought we need to speak was when the rates were rising very quickly. We were screaming into the void of like, no, no, no, this business is fantastic. At ZIRP, it is fantastic. At 5% fed funds rate, it is just fine. At a number that's higher than that. We love our value creation opportunities more or less at any rate, like obviously, you know, it breaks at some point, but US Economy breaks at some point for all possible values of the federal funds rate. We're going to be just fine. And a long explanation why and I don't think anybody heard it, people like, yeah, yeah, but, but the rates are up. So you're obviously.

[Host 2]

Yeah.

[Max Levchin]

And so the dip in return of our stock is like, oh, a great turnaround, but it basically amounts to people saying, wait a second, this thing is in just as much demand, does just as well, grows just as fast, just as profitably through whatever rate cycle.

[Host 2]

Yeah. I think we've talked about the history of the effect of interest rates on a firm. But we're now in a new regime and the fear is not higher rates, it's higher inflation. And so do you already have an intuitive sense of how a different inflation regime or different inflation environment will affect a firm? Are you confident? How do you think about the relationship between.

[Host 1]

Yeah, it's like I can imagine somewhat of a double edged sword and that like higher prices means people are more likely to want to use credit. But maybe there's a drop in some purchasing activity of some items also.

[Host 2]

Just like if you're getting $10 in two years and it's worth less, that should have an effect. But how is it actually playing out?

[Max Levchin]

So a little Too early to tell, but you are exactly right in a sense that we saw this game play out in the last inflation spike just a few years ago. We absolutely saw an increase in demand because as things became a little bit more expensive, people felt that it would be better if they weren't paying for them upfront. People who initially would say, yeah, it doesn't really make sense for me to finance this thing. I have the cash like, ah, I kind of want my cash to go a bit longer because I don't know exactly which way the prices are going to go. So we expect more demand. We have to be, we're always very careful. The number one line I give at every investor event is credit is job number zero. As a computer science major, I count from zero and it is the most important job. We grow as fast as credit performance will allow us to grow. Full stop. We have to print consistent credit returns otherwise we lose credibility with our debt investors. And that's the most important thing. So we will grow fast, no faster than credit results will allow us to. We expect more demand. We don't yet see literally anything related to changing credit performance. We just printed our results and they're just as we expected them.

[Host 2]

Yeah.

[Max Levchin]

And so TBD whether we see some deterioration of consumer credit due to prices, we really didn't see much of it at all through the last inflation cycle, which was quite dramatic if you remember. And so we feel pretty great about our ability to underwrite kind of all eventualities. We are obviously not blind to the macro reality. The minute before I got on stage and promised people 25% growth with an increased profit margin. So we must think we're going to be okay. And, and that is in fact the case.

[Host 2]

Yeah.

[Host 1]

How is AI changing how you think about international expansion? Does it make teams more efficient in terms of adapting the product for different markets? Everything from language to just simply being able to ship more code? Or is your international GTM just much more like we know what it takes to launch a new market. If AI makes it slightly more efficient, great. But it's not going to change the strategy.

[Max Levchin]

I think the 1A of every list of things that have really changed thanks to AI is shipping code. I think this is the best time to be a CEO with a computer science degree because you feel what this means. You know what it's like to make code. You've done it in my case for the last 40 years and I know how much easier it is, how much more effective teams can be, how you can have three people in a room Build a product over the weekend, which is kind of an amazing thing that didn't exist six months ago. And so shipping code is absolutely. We're seeing that in. My latest shareholder letter actually showed an illustration of percentage of code written by AI at a firm. And it's like, it's not like a slow rise. You're like, okay, we're trying, we're trying. We're convinced 60% to start, 75% last month. So it's just rocketing. It's not like people are writing less code or reviewing less code. It's just that much more productive. So that helps us internationally, helps us domestically, helps us with everything. I think other parts of the business are much better with AI. So translation, obviously services are fantastically effective. With AI you can do a lot of interesting things. With legal, anything that's text heavy, you have just a great helping hand. But. But the sort of night and day moment is writing code that's 1/8 or maybe like 1p on the list of things that just stunningly effective.

[Host 2]

Are you expecting token cost to be like an important line item in the P and L going forward? Like we've seen? Because there's so many stories of a lot of code written by AI. Sometimes that moves the needle, sometimes you're just sort of doing what you needed to do and doing it more efficiently or faster. But at the same time, if you get hit with like $1 billion token bill in a year, that might be material. And so you start having to do. And it feels like we're at the early innings of ROI calculations. We're more in like the feeling of like, oh, everyone loves this. We're seeing good results.

[Host 1]

Max doesn't strike me as someone who's just going to tell your team to slap it up a lot of AI because obviously there's stories coming out of Amazon and Meta where teams are basically just like, I want the optics to be that I'm using more AI than anyone else on my team, so I'm just going to run things overnight for no reason.

[Max Levchin]

Yeah, we are extremely, extremely metrics driven company. Some might take issue with just how completely obsessed we are with you. Improve what you can measure. If you're going on fields, you're going to, you know, feel great until you don't. So we measure everything obsessively. We have at the moment, at least fantastic ROI on our token spend. We are, I would say, you know, it's real numbers, so it's not like, oh, you know, don't care, don't look, we have a weekly report that's generated thanks to AI and some very smart humans who are monitoring our token use and tell us how each team is producing value and what it's costing us in tokens and all the sort of various conclusions downstream. So we are already very mindful of what it means in terms of return on investment on tokens, return to shareholders at the limit. Right now it's undoubtedly a very, very accretive thing, but we're also not assuming that infinite budgets are just great. And we also expect, by the way, prices of tokens will probably increase over time. I think there's plenty of subsidies taking place. You're now inside one of those places, so you probably know the true

[Richard Socher]

if

[Host 2]

we can circle back to specialization from AI creating effectively more generalists. I'm interested in computer science history, your career history. How have you processed? There was a point where front end and back end engineering were two very separate disciplines. At the point where you could do JavaScript on the front end and the back end, you see more full stack engineers. Now you have engineers that are designers, designers that are engineers, designers that might be pushing server code and stuff. So I'm wondering how you're thinking about maintaining that idea of you want the specialized genius, the artisan, the true expert in your organization versus you also want to empower someone to move really fast with a small team. And so you might not need to staff 100 experts to do one thing. And there's got to be a tension there. So I'm wondering how you're grappling with that.

[Max Levchin]

It's a great, great question, and you're totally right. So there are definitely pockets of today's software engineering that accrete profoundly to specialists. So if you are an extraordinary infrastructure engineer, you're far from being replaced by some mindless drone because it's as much art as it is science, intuition, experience. I've seen this kind of a failure before. AI is good at it, but AI is a great tool. You're in no danger of being supplanted just because so much depth goes into that. And by the way, AI making a slight mistake could cost you a massive outage in the case of and so the cost of error is very high. Like an even better example actually in my world would be an underwriting model. So somebody who designs underwriting models like, hey AI, go build me a great underwriting model. Come back when you're done. Like, here it is. I hallucinated one for you. Like that's like real money lost that we will Never see again if it in fact didn't hallucinate the right thing. And so many of these sort of ultra specialist roles, they are absolutely benefiting from AI. But that human is not just in control, they're not just steering, they're like reviewing every line of code. They have another human making sure that there's absolutely nothing that could go wrong. So that's sort of one thing. The completely other spectrum of this, which is I think what you're alluding to. So, you know, fun examples. We just had a product hackathon, so only product managers can attend. A lot of our product managers have CS degrees, so we're cheating here a little bit. But these people were like, hey, I love writing code, but I want to make it into things as opposed to code and like I want to ship a full product. So most of those folks haven't written code professionally in quite some time, maybe since college. They're really good with figma. They know how to design things, they have a sense for taste, but they don't remember anymore what it's like to go deep into some SQL query debugging. We had 100 people, 37 projects in 27 hours went from a whiteboard plant to a shippable feature that was presented to the entire company, like, hey, we have this thing, it's. And we graded it not just on. So the winners and the whole grading of who won 1st through whatever, 10th were all people that had to report. Not just like, here's my idea, here's what it looks like, here's my prototype. How close are we to shipping it? Are you ready? And the winner was absolutely, probably the best idea, but also the ones that said we're ready, like if we're allowed to ship it tomorrow, we will. So that is like an extreme generalist who has a vision and few tokens and a feature in 27 hours with like two of their co conspirators in like a pizza box. So that is new and different and it's totally a new phenomenon.

[Host 2]

This AI thing is going to be terrible for the pizza industry because famously two pizza team, you only need one pizza teams now.

[Max Levchin]

50% reduction.

[Host 2]

50% reduction, short dominoes. This is a disaster. I hadn't thought about this.

[Max Levchin]

We sprung for a somewhat higher quality.

[Host 2]

Oh yes. Okay. So revenues, flat margins, potentially going up in the pizza industry.

[Host 1]

Okay, last question. How are you? How are you? What's your framework for evaluating the. The all the different AI products being pitched to you across the rest of the organization? You have the benefit of being A technical, you know, CEO, a lot of other CEOs in your role or just like, I just want to buy AI everywhere. Just give me every part of my company. Just give me, give me, give me three pilots.

[Tyler]

I'll pilot anything.

[Host 1]

Whereas I think you're probably looking at some of this more non, you know, non deterministic work saying like, hey, you know, you can play around with it, but maybe you don't have budget.

[Max Levchin]

So I've a great heuristic. This, this will be potentially worth, worth the rest of my drill. If you know how to describe the evaluation criteria being used by the maker of the tool you're buying, it's probably worth piloting. If you can't, it is slop and you're being sold a story. We know Codegen works and is amazing and it's only getting more amazing because we know how to validate code. And you can use another AI to validate code, but you can also just look at does it work? Does it pass the unit test? So the feedback loop of I gave you a thing like codex was okay, and then it was better, and then it was great and now it's really pretty damn good. It's not because someone was coding faster, it's because it has really objective eval criteria and it just loops on making itself better with some degree of human influence. The same can be said for a bunch of industries, it cannot be said for others. I'm not going to dig into examples too quickly, given we're short on time. But if you can say, oh yeah, I know exactly what these guys are doing. Customer service centers. Is the consumer happy after they hung up the phone with an agent? Can we ask them? Yes.

[Host 1]

What's the satisfaction rate?

[Max Levchin]

Is there a way to say that was good, that was bad AI do more of the good kind, Fast feedback loop, high quality eval criteria. So we use a ton of agentic customer service because we know it'll just get better and better and better. That's a great thing to buy. We're not a specialist. We are excited to buy it from the specialist. Somebody tells you, oh, it's AI is going to be a tool that's going to make you a better writer of fiction.

[Delian Aspro]

Yeah.

[Host 1]

Or even pitch decks are an example because like you could have the worst formatted pitch deck and have having tremendous success with it. But it didn't have anything to do with the deck. It had to do with your delivery and who you are.

[Max Levchin]

And so I would not, would not buy pitch deck AI assistant, even if somebody paid Me to do it because it's 95% the talk track and the company so far and the company and the idea and the market and the TAM and all those things and so. But most importantly, 100 companies pitch their decks. Good old and made by AI. Yeah. 7 out of 100 get funded. 93% failure rate sucks. Should never raise money like. Is that the conclusion? So I would stay away from the non objective or very slow feedback loop systems before I consider buying tools there.

[Host 2]

I love it. Well, thank you so much for taking the time to come chat with us.

[Host 1]

Great to catch up. Have a great week on the progress.

[Max Levchin]

Thank you. And I'll come back with smelting news.

[Host 2]

We'll talk to you soon. Have a good one. Goodbye. We have deli announcement Brujo from founders Fund embargo joining in just a few minutes while we wait for him to get here. He's coming. In person we will talk about Joe Low, who I thought was a household name but apparently is not. He is the fugitive behind the 1mdb scandal. It reads like IMDb it's not 1mdb. It was a Malaysian financial fraud that led to the disappearance of $4.5 billion. And he's in the news today because he asked Donald Trump for a pardon while on the run. I believe so. JHO Low, the alleged mastermind of one of the greatest financial frauds in history. You gotta read this book. The camera's following you around.

[Host 1]

That's the new meta.

[Host 2]

New meta. Go on the run until you can get the part.

[Host 1]

Go on the run and you're just popping up to say, I'd like a pardon, please.

[Host 2]

Yeah, yeah, yeah, yeah. So, yeah, you gotta read this book. Billion Dollar Whale. Fantastic book. I thought it was gonna be turned into a movie at some point. It should be. It's a fascinating story, but we'll just give you the high level and we'll do a whole deep dive later. Joe Lowe, the alleged mastermind.

[Host 1]

We should take matters into our own hands. Make the movie into a play.

[Host 2]

A play. Okay. A stage play. I like a stage play.

[Host 1]

We haven't done a play yet on set.

[Host 2]

No, not yet. Although a lot of the red string poster board sessions, they feel like plays. They're very well acted.

[Host 1]

But I'm telling you, we're reading off.

[Host 2]

Yeah, yeah. We don't have lines for those.

[Host 1]

Joe Lie one pardon into the room. Billionaire Lee.

[Host 2]

Yeah. So Joe Low, the alleged mastermind. Alleged mastermind of one of the greatest financial frauds in history is asking President Trump for a pardon. The Request was filed in recent weeks, according to people familiar with the matter, and if granted would remove us criminal charges against the fugitive Malaysian financier. A Justice Department website lists a pending request for pardon for completion of sentence under Joe Low. The move represents the latest gambit in the extraordinary saga of Low, a once little known business businessman who prosecutors, prosecutors say used subterfuge, fake documents and payoffs to engineer the heist of billions of dollars from One mdb, which was a Malaysian government owned investment vehicle set up to promote economic growth. He's known for his lavish lifestyle, partying with Hollywood stars and political rulers. He was involved in. There were a bunch of other people that went to jail for this, but he financed the Wolf of Wall street in potentially the most ironic thing you could do as a fraud, finance a movie about a fraud.

[Host 1]

So it's illegal to make awesome movies now.

[Host 2]

I guess it is a great movie and you can tell that the budget was really flying with that one because they crashed a Lamborghini Countach which is maybe $500,000 car and they just, I think they just destroyed it and they destroyed maybe the actual. I don't know, that might be. Not real, but very fun movie. Anyway, we have Deli and Asproo Hubbard from Varda and Founders Fund in the TVP and Ultra dome.

[Host 1]

There he goes.

[Host 2]

There we go. Have a seat. Oh, he's doing a victory lap. He's doing a baby laugh. We love a victory laugh running around the table. For anyone who's been living under a data center. Introduce yourself.

[Delian Aspro]

Hi, my name is Dellene, co founder and president of Varda. Have a couple other side gigs, but that's the main game.

[Host 2]

Let's focus on Varda today. Oh yeah, we're talking space data centers, right?

[Delian Aspro]

We do have some chips up there. So in theory you could do. But we do more valuable things than just some inference. We make some drugs up there.

[Host 2]

And there's a big announcement today.

[Delian Aspro]

Yeah. So for the first time in human history, there is a large publicly traded company, United Therapeutics. $25 billion in market cap traded on the NASDAQ. They're using their own balance sheet to go produce a product, physical product, in low earth orbit with us pharmaceuticals. They focus on pulmonary disease. So think lung blood, et cetera. We're working on those products together, crystallizing them into a better formulation, bringing them back down. That's the only company that makes physical products in space. Everybody else, it's all radio waves, it's Internet inference. It's photos of the earth et cetera this is literally the first time in human history where there is physical products being moved in space as a way

[Host 2]

to generate revenue there or real business.

[Host 1]

And that's happening, or is about to be happening.

[Delian Aspro]

Yeah. So we signed this deal. You know, call it, like, you know,

[Host 1]

stop you right there. Because there's an opportunity for someone else to put a. Oh, it's a race and a balloon. We should. We should race. We should race a big.

[Richard Socher]

No, no, wait.

[Host 2]

Balloons can't go to space.

[Delian Aspro]

Balloons can't go to space. You got to go. You got to go on the rocket.

[Host 1]

A balloon. And then once you're up there, jetpack type thing to.

[Delian Aspro]

Yeah, people have talked about that, like, you know, could you go float a rocket up there? But it turns out the, like, size of planet, balloon and complexity, that you're just better off building a rocket to get the rocket up there.

[Host 2]

Yeah, I talked about a space ride. They would. They would take the balloon up and. Yeah, and this was actually used back in like the 30s, I think.

[Delian Aspro]

I think they did some, like, very early tests when rockets were small. Yeah, yeah. Now the rockets are big and like, satellites are big.

[Host 2]

A little.

[Delian Aspro]

A little harder to get up there. But, yeah, we signed the deal a little while ago. We've been working on the products with them for a period of time in order to get ready to go to space. We do do some work on the ground to understand, hey, what is the thing that we're going to go do in space? And let's make sure we figure out all the other process parameters. When you're like crystallizing a drug, there's a particular temperature, a mixing rate, a solvent. We just bring one extra knob, the gravity knob. Now, here on Earth, we can't decrease gravity, but we can increase it. And so alongside the customer, we look at basically adjusting that gravity knob down here on Earth, figure out kind of how to tune all the other things. So we're tuning the right knob. Once we get that all set, load that up into a spacecraft, send it up to space. Obviously there, the gravity knob goes to zero, get a different result, land it in Australia and then send it off to the clinic. We're still working with United on the ground so far, getting those drugs ready for flight. But we'd expect basically next year to fly the first United base.

[Host 1]

And how are they making this drug on Earth today?

[Delian Aspro]

Today you're using traditional techniques.

[Host 1]

So there's a whole set of companies much less efficient.

[Delian Aspro]

You're just getting a different result. So as an example, the sort of use case or paper that the company was in some ways predicated on. Obviously there's been 40 to 50 years of work on protein crystallization up on the international space station, Skylab space shuttle. But in 2019, Merck basically showed their blockbuster cancer drug Keytruda. They crystallized it up in space. They showed they could shift the administration from what was previously an IV drip to now a subcutaneous syringe. So you don't have to go to an IV clinic. A rural patient that can't access IV clinics can now access this drug. And so that's what we're doing up there with United Therapeutics is we're looking for basically changes in performance of the drug because of those different results that allows them to shift administration. United Therapeutics is famous for a variety of different things. So the company was originally started by Martine Rothblatt to solve a rare pulmonary disorder for her daughter. Built that up into a huge platform. They've really leaned into the frontier over the course of the last year, so 20, 30 years including. They make gene modifications for pigs to create human organs inside of pigs. But one of the things that they're known for is for one of their pulmonary compounds. They shifted the administration from a nice big sound. Well done, well done. I like it, you know, a lot better than waiting on like a, you know, whatever 24 year old kid to die on a motorcycle. Your kidney instead, just get it from a pig farm. But one of the other things that they're known for is they actually shifted one of their sort of primary drugs from an oral pill to instead of inhalable and made it much more efficacious, easier for the patient because you have to take the drug less often. It's things like that that we're looking to sort of do with them.

[Host 2]

Dumb question. You, you have the gravity knob, you're able to turn that off, but aren't you adding a knob for vibrations and basically dialing it to 10? Does that ever affect. Because you imagine you're sort of centrifuging it like it's gyrating as it's re entering. So is that a problem? Is that for the selection criteria, do you have mitigation strategies?

[Delian Aspro]

What you can think of is like in a pharmaceutical process where you're crystallizing something, you're basically going through like a phase change, you're going through from a liquid to a solid.

[Host 2]

Okay.

[Delian Aspro]

Where things matter is when that phase change is happening.

[Host 2]

Okay.

[Delian Aspro]

In that particular moment is where you don't want the sedimentation, convection Et cetera. So we do in some ways, we like load up all the liquids ahead of time. Those liquids for sure on the rocket, light it up, get shaken, et cetera, but it doesn't matter because the liquid's still saying a liquid.

[Host 2]

Yeah.

[Delian Aspro]

Then up in space, what we do is we go through that phase change. We go from basically like a liquid to a solid.

[Host 2]

Okay.

[Delian Aspro]

That needs to happen in a very pure environment. We shut down all the reaction wheels. We make sure that everything on the spacecraft that would vibrate is totally silent. So it is as pure of microgravity basically as possible. It then becomes a solid. And at the end of it, it kind of looks like a salt, basically. Like think of it like table salt. Now on the way back, that table salt for sure is shaking a bunch. But the table salt, if you think about like, you know, some table salt that we put, and we go put it in like a centrifuge for 100 years at 1000 GS, it's still going to be table salt.

[Host 1]

Where are those, like vibration heat more

[Delian Aspro]

you do need to manage heat. If you melt the salt, you're fucked because it's back to a liquid. But, you know, this is, is what convinced United that we would be capable as a partner is we actually had data sets from our very first mission where the surface of the heat shield was hotter than the temperature of the surface of the sun. So like extremely hot on board. I think it was like something like 14 degrees centigrade, plus or minus 0.1 degrees. And so we had extremely precise temperature control on board. And so a lot of what has brought people to the table for this sort of first commercial deal is while there's been this historical work on the iss, nobody had showed how do you go do this on a purely private platform, fully autonomous, low cost, high cadence. And a part of what convinced them was, yes, both the regular flights we've been doing, right, we are now on w, sort of 6, 7, 8, 9, will be sort of launching later this year. But it was also, hey, not only are they launching and landing these things, but on board they've got the right process control. So I can see, hey, they can go work on my drug. And by next year, when we fly seven times in a year, that's going to be higher cadence than everything that the ISS, space shuttle, etc. Have done in all of these human history, let alone in 2028, 2029, as we continue to go higher and higher cadence.

[Host 1]

So how many things need to go right in a row? To have a successful mission, I imagine that, you know, there's like all these

[Delian Aspro]

different, like just the space mission or like get the drug into the clinic, because the drug in the clinic is more complex.

[Host 1]

Just the space mission.

[Delian Aspro]

The space mission. At this point we've like kind of shown we can do pretty safely. But I'll walk you through it. First on the ground we gotta go build basically effectively think of it as like a two part spacecraft. One is sort of looks like a traditional satellite system. It's a little bit lower power and under spec'd relative to most satellites because we don't need super high power to communicate with Internet terminals, radio waves of photos, et cetera.

[Host 2]

Oh, that makes sense, yeah.

[Delian Aspro]

The only power we need is just to run the bioreactor and the bioreactor doesn't consume that much power. So think of it as like it

[Host 2]

is like basic telemetry.

[Delian Aspro]

Yeah, basic telemetry, but we don't need like super high pointing accuracy while it's just floating up there. Cause again we're not pointing at a particular place on the ground to basically try and take a photo of it. And so think of it as like we build an under spec'd sort of satellite. Now the one thing that it is big on is like it's got a lot of thrust and propulsion on board because we need to come back unlike every other satellite, but otherwise under spec satellites. So we do those on the ground. We've now built a handful of those. Then we have to build the second part of the spacecraft which is the re entry pod. All of this does have a relatively complex supply chain. Think like the solar panels, the GPS chip, the flight computer, that type of stuff we buy out of house. Some things you have to do in house, you can't buy a capsule like off the shelf. So we have to build the heat shield, the structure, et cetera, around that. Those have to get integrated. We have to do a bunch of ground tests to make sure that works. Then you ship it off to SpaceX's facility a couple months before launch. Make sure that everything's still working over there, install it onto the rocket. Once we're on the rocket, we kind of look like any other satellite that goes up there. Whether it's like you know, asteroid mining satellites or comm satellites or spy satellites. SpaceX takes us up to orbit, they then drop us off again like any other satellite. We spend between, call it like 4 to 12 weeks up there. Kind of depends on the drug manufacturing process. Some go pretty quick, some take quite a bit of time. While we're up there, we have basically pharmaceutical scientists on the ground in Mission Control. So you remember those like Apollo films in Mission Control we literally have like think like Amgen scientists, but like they're in mission Control talking to the bioreactor.

[Host 1]

I can imagine you love hanging out in there.

[Delian Aspro]

Brui always makes fun of me. He's like, you know, I don't quite know the American idiom or don't get it, but hopefully you guys will. But he's like, he calls me like the gentleman's mission control where there's like the gentleman's farmer. It's like the guy that stands around the farm but it's like obviously not really doing anything. I'm like the gentleman's mission control where like yeah, I like hanging out there. Obviously I like, I'm completely useless and mostly a distraction. So I try and mostly just listen into the live stream now.

[Host 1]

Every team needs a class clown.

[Delian Aspro]

Yeah, exactly.

[Host 2]

Can you unpack a little bit more about the orbit? There's a piece in the Journal today. Space data centers face a lot of challenges. They all gotta go to sun synchronous orbit. I imagine you don't, but I want to know about the knock on effects of like, what can you tell me about sun synchronous orbit? What orbit do you use and why? And then how might this change what happens in the future?

[Delian Aspro]

Yeah, part of what's beneficial about sun synchronous orbit is you're basically like always like facing towards the sun and have sunshine. We are not as dependent on that as I mentioned earlier. We're not super power constrained and so we can just charge up, you know, basically our batteries on board. And while we're on the dark side of the Earth, we can basically just drain those batteries, recharge them basically on like the next orbit. We're up there just for the microgravity and so we can be in almost any orbit. That gives us a ton of benefits. One, when you're in space you don't have to worry as much about traffic. We just go where other people aren't. But two, it actually helps a ton with launch. You know, as an example, later this year for the first time we'll be actually launching. Historically we've only launched to sun synchronous orbit because that's where all the transporter missions, which are the SpaceX rideshare missions.

[Host 2]

That's what I was going to ask.

[Delian Aspro]

They now have what are called bandwagon missions which go up to. They're also rideshare but they go to Polar orbit. Yeah, we'll actually be flying on our first Bandwagon mission later this year. So now we have launch flexibility. There's a bunch of companies that can't go on bandwagon or basically we're actually probably one of the only companies that can go on both. Either.

[Host 2]

Yeah, that's what I was going to ask. Because what happens if everything winds up being Sun Synchronous orbit? Because I think there's 10,000 or so Starlink satellites. Elon's asking for a million for space data centers. That means like 90% of the launches are going to be Sun Synchronous. But that's okay for you?

[Delian Aspro]

Yeah, we're okay going to Sun Synchronous. We can quickly get out of the orbit if we need to try and go so slightly off of it, or we can just go to Polar and be totally sort of happy there. And that gives us some benefits on the launch economic side of things, where now we're basically able to go on any rocket.

[Host 1]

Yeah. So I was going to ask. It seems like every hyperscaler, some startups are in, some of the labs are thinking about space. Are you expecting like launch to could, you know, theoretically increase. Could be like bottleneck launch.

[Host 2]

It's been falling for so long, but yeah. What do you expect?

[Host 1]

Yeah. How are you planning? Because you have to buy capacity. Launch capacity years out.

[Delian Aspro]

Yeah, yeah. To date launch prices have only gone up. So they have not gone down since we started the company. Launch costs for SpaceX, I'm sure have, but there has not yet been pricing competition. So why would they necessarily pass on those savings to be. And they haven't, I think, seen that much price elasticity in the market. When they've like lowered prices, it's not like they've seen like 10x greater demand. And so for them it's like capture as many margin dollars basically as possible.

[Host 2]

Yeah.

[Delian Aspro]

I do think what we're going to start to see is more competition in the launch market over the course of the next five years rather than the prior five. Right. Blue Origin, I think it was today or yesterday, announced they're going to be doing their first external sort of fundraising, which speaks to, you know, I think them taking themselves much more seriously as a company that is going to like regularly show shareholder value and launch very regularly. You've got Rocket Lab now. It's like a, I think $65 billion company. Yeah.

[Host 2]

Wow.

[Delian Aspro]

Definitely have enough resources to go tackle Neutron and they've got shareholder expectations for them to get that medium lift, you know, sort of rocket up, you've got stokespace. And so there's a few others basically playing in it. We are booking out, you know, sort of rocket launches now all the way through 2029. So we do have the capacity that we need for the next year, sort of many years. And I think by the time we're getting into booking late 2029 flights. Yeah, there's a world where there start to be basically more providers online. SpaceX, at least, basically to date has been like, for us, at least as many times as we've wanted to launch, that capacity has been available. They are not capacity constrained, they are still more demand constrained. I do think that because no one's

[Host 1]

making anything in space, totally nobody's making it.

[Host 2]

And that's why they were able to be so successful with Starlink, was because they had the residual capability.

[Host 1]

Right. It's like tealism. Right. It's like, like we, we created space, we created the space economy and then it's just more Internet.

[Delian Aspro]

Yeah.

[Host 1]

And now it's just going to be more, more AI. How are you thinking about other drug categories? I can imagine Space Retta would be a pretty hot drug if people would smoke space, rather.

[Delian Aspro]

Yeah, one, one thing that I didn't realize is up so Ozempic today, one kilo of Ozempic retails for a million dollars. So what people don't realize is when people ask me like, oh, space drugs, economics, et cetera, does this stuff work? I have to tell them, like, yeah, for you it might be like a thousand bucks basically per shot, but it's because there's very, very little Ozempic basically on board.

[Host 2]

A lot of them are like powders that then get diluted and then injected. And so, yeah, the actual core ingredient's so tiny.

[Delian Aspro]

Exactly. And what we do up there is we just make the powder.

[Host 2]

Yeah, exactly.

[Delian Aspro]

United has United Therapeutics, our partner that we're announcing today. They have a ton of expertise, expertise in all things like, you know, rare pulmonary, certain disorders. They're for sure going to be the person that we work with on that indication, you know, scenario. But microgravity has a wide set of pharmaceutical applications. Neurological, immunology, oncology, et cetera. And so for sure we'll start to, you know, I think in the success case of Varda, it's not going to be like we have like 50 different partners. It's probably going to be like five to seven that we go really deep with. We kind of choose sort of one partner per your sort of indication. Area that we go as deep as possible. You know, ophthalmology is another area that we're looking, you know, so deeply at, you know, partially because like the eye, smallest organ, again, smallest amount of powder that you need. Also an area where some of the other changes of administration technologies don't really work as well. And so, yeah, we're starting to explore who are the partners for the other indication areas.

[Host 2]

What about beyond bio? I remember ZB Land being something that people were talking about potentially being able to manufacture microgravity. Wildly different economic trade off there. But are there other, when you imagine like the industrial city in Leo, like are there other categories that extend maybe 10 years out, 20 years out where it might be logical?

[Delian Aspro]

Yeah, I think like if we look at the next decade, you know, call it, if there's 200 products that are manufactured in orbit, I think 195 of those are going to be pharmaceuticals. So vast majority is going to be sort of there. And that's where our core focus area is. If I had to sort of peer into the crystal ball and look at what comes next, I do actually think probably fiber before semis is probably sort of my rough guess. Yeah, when we were looking at that like high end, you know, sort of Z Bland, you know, fiber market back in 2021, when we started the company, there was clearly some level of like, you know, interesting capabilities there, but just like relative to what we were seeing on hypersonic pharma, not as large. Interestingly, that market has matured a lot. There's sort of two sort of call three areas where it has applications. One is it is the lowest. Basically, you know, think of it as like interference inside the optical fiber for infrared. Near infrared is now used in a lot of the high energy weapon systems. So if you think about having to pump that laser through a fiber, you basically want as little basically resistance as possible. There's just a lot more being spent on those types of near infrared weapon systems, especially with some of the anti drone stuff that's coming down the pipe. So that's sort of expanded. It's also used now a lot more in metal cutting and organ eye surgery relative to where it was sort of five years ago. So that's expanded. The other area is quantum communications. So quantum entangled photons are much more sensitive to those types of impurities in fiber. Quantum comms has also grown a lot over the last five years. And then the other area is oddly data centers. You know, I was going to say

[Host 2]

as soon as there's a data center story there's going to be 25 companies funding IPOs and stuff.

[Delian Aspro]

So ground data centers, not space ones, but they are actually.

[Host 2]

No, no, no, no, no. You make the ZBLAN there, you bring it back and then you.

[Delian Aspro]

Dude, zblan is now being used in terrestrial data centers. That was not a use case at all five years ago.

[Host 2]

So all these things and optical.

[Delian Aspro]

Yeah, they're still like early I would say. But in like, you know what's happening though is also like the VARD economics keep improving on a flight by flight basis. The Z bland market keeps growing. Yeah, we're like for the first time, I would say we're like dedicating resources or anything. But like around the office, for the first time in five years somebody was like the Z blade stuff is like almost starting to pencil out again. It's probably like four years till we like, you know, sort of do anything. But like you're starting to see those lines start to get closer, closer together. That's what I love about a platform like ours is like if you had told me like predict where your revenues would be five years ago, really hard even this year. Like things that are landing are probably not what I would have underwritten a year ago. Like I think pharma has like way exceeded expectations. Like it's pulled in probably like three and a half years relative to like, you know, our series C that we announced last year. Our financial projections on that on pharma were just like totally, you know, sort of wrong. And so I'm excited by the idea that like, yeah, there might be other areas that are wrong. Semis I think is just going to be like, you know, the furthest out partially. Just like the process equipment and the sensitivities. Like if you think about a bioreactor, it's like hard, but it's kind of like mixing like liquid together. The hard thing is like which liquids. That's what's hard, the mixing of it. Not so hard.

[Host 1]

Yeah, but you're making those decisions down

[Delian Aspro]

here on the ground. Fiber, same thing. It's like, you know, it's like once you like make that like precursor like rod, then drawing it is not like a crazy complex process. It's making that rod that's hard. But again you're doing that on the ground with semis. It's like, oh my God. That process like so, so sensitive. Right. If you look at like the like stuff that they use for euv basically like lithography, the mirrors for that. If you were to lay out a lithography mirror over the course of the size of Germany, the amount of variation you see in terms of mountains would still be less than 0.1 millimeters. And so this stuff just has to be so, so, so, so, so, so, so precise that it's just like the

[Host 1]

cost is so high.

[Delian Aspro]

Yeah.

[Host 1]

Even during the experimentation process you could.

[Host 2]

Yeah. The famous quote from TSMC is that if there's an earthquake, they don't have like a, a pager duty. Like they don't slack everyone, hey, you got to come in, there's an earthquake. Everyone knows if there's an earthquake, just go to the factory.

[Delian Aspro]

Exactly.

[Host 2]

And the employees just know because it's just so, so important.

[Delian Aspro]

And so just that like microgravity process equipment is just so much more complex. So it'll get there one day. But like bioreactors, the fiber draw things are like orders and orders of magnitude simpler to make. Like microgravity bioreactors, microgravity fiber draw towers. Yeah, yeah.

[Host 1]

Will it ever make sense to have a permanent like ISS style structure that you are transporting supplies to and then sending, you know, finished product down? Because like, knowing you like you would like to have a proper like dedicated space factory. But does it actually make sense?

[Delian Aspro]

It's always been the somewhat secret, somewhat not secret, like ambition of Varda is to go build that, but go do it step by step.

[Tyler]

Right.

[Delian Aspro]

So when you look at like generation one of basically what we're building, it's this two part spacecraft, right. Satellite pod. Satellite has all the process equipment, et cetera. Pod just has the finished good. When we're done, pod survives all that process equipment, everything on board. We basically just burned it up. Generation 2 that'll be flying in like Hall 2029. Think of it as just like we make it so that that heat shield material just envelops everything. And so now the satellite plus the pod, they both survive. And that'll probably look like a little mini space plane is kind of how you can think about it. So space plane has the solar panels, the propulsion, it's got the process process equipment. It's also got the finished good that goes up and down. And so now that full system is reusable. Improves the economics. We'll have more payload on board too. Now at some point it doesn't make economic sense to keep moving that process equipment up, down, up, down every time call. By the end of the decade, the goal will be we'll start to invest into those like fixed pieces of infrastructure in orbit. Early on they'll probably just look like a satellite just with a Little docking port. Basically the like space plane will come with the raw ingredients, basically the liquids, etc. It'll do dock with that like satellite up in orbit, exchange for the fabricated goods. But then we'll have like 10 of those satellites. We're start to stitch them together. Basically looks like a station. One day there'll be so many of those pieces of equipment on board that we'll be able to economically justify, hey, probably need to do with a wrench every once in a while to go like fix some stuff up there. But like once we can economically justify like one person with a wrench, 10, 100, 1000 and so like yeah, we for sure we want to build the first industrial city in low earth orbit, but we're taking it like one step at a time, building out the business, getting great unit economics, proving out one use case at a time.

[Host 1]

What about the other parts of the stack that you actually need for that? So like I'm thinking like docking. Is there a startup that's just, just working on docking?

[Delian Aspro]

Yeah, I mean, you know, sort of credit to them. There's a company in Seattle called Starfish, you know, sort of space that is hyper focused just on like the like docking, maneuvering, the software for that, that

[Host 1]

feels like something that like if that's not your core competency, seems like you could have a series of nightmares around just trying to figure out docking. And that is a probably super widely applicable process for the rest of the space economy.

[Delian Aspro]

Totally. It's like back in the day we would have had to make our own solar panels, our own thrusters, propulsion tanks, et cetera, for the stuff that we build on the ground. That stuff over the course of the last decade has become a much more mature supply chain. As we start to get to this space station world, I think there will be some of these things where Starfish I believe, basically makes this camera that you can basically attach to your satellite. They take over the controls for your propulsion and they'll dock you to wherever you're needing to go. Yeah, five, ten years ago, there's no way that we would have been able to buy that. We would have absolutely had to build it. And so yeah, I think on a ongoing basis we'll always be doing that sort of build, buy, trade. And ideally, obviously if there's an off the shelf solution, integrating that and that only accelerates basically our progress. And so yeah, it's cool to see that as obviously Varda is developing the general space economy and all these use cases are developing and heck, when you said like you know, sort of warehouse in space. There are companies like Gravitics that are trying to do like dedicated warehouse vehicles. Maybe there's some world where like yeah, we like buy the warehouse and we just install some docking ports, fill it up with process equipment. But we don't have to like build the warehouse. I'm not like opposed to that. What we really want to focus on is like we're have the best lab down here on earth to like assess how to make these formulations. We definitely have to make the bioreactors ourselves. Nobody else is going to make micro gravity bioreactors. We got to make those space planes basically ourselves because, because nobody else is going to be bringing those up and down all the time. But the rest of it pretty open to buying where we can.

[Host 2]

How are you thinking about the public markets? SpaceX is going out in the trillions. A lot of bio companies, I'm sure United Therapeutics went out in the billions probably. And a lot of biotech companies, they go out earlier, pharmaceutical companies go out earlier. The public market, sometimes they like really stable projections. Do we understand the quarterly forecast? Sometimes they like a big vision and you get out there with a Tesla or something and it does really well. What is your take on how you might fit into the public markets in the future?

[Delian Aspro]

Yeah, I mean there's not a lot of private like 10 billion dollar in market cap like pharmaceutical companies, they basically all go public before that.

[Host 2]

It's like, I mean there's so many deals like every day a big biotech company is buying something for 1 billion, 2 billion all day long.

[Delian Aspro]

Totally. That's basically the default.

[Host 2]

Yeah.

[Delian Aspro]

Now I think like GLP have shown that like in the therapeutics world you can build a like you know, business that can effectively be like a trillion dollar outcome one therapeutics business. You know we're definitely thinking about that from the Founders Fund. Hat side of things, how do we back things that aren't just that one to $5 billion therapeutics outcome? Edvard I think we'll look at it like on a round by round basis where like I think about it as like cost of capital. The public markets are offering me a lower cost of capital because they have more expertise because there's a bunch of hedge funds that have like MD PhDs and tons of them on staff to deeply understand pulmonary diseases, ophthalmology, et cetera. Yeah, we'll go to the public markets if that's the case. If we're still seeing tons of support in the private market and we think it's like relatively not similar, you know, we'll study private for a period of time. If I had to again, peer into the crystal ball, my guess is like, yes, relative to a SpaceX or an Anduril or things like that, I think we will have much lower cost of capital early on, you know, sort of being public rather than not. Because the public markets have so much appreciation for the types of clinical data sets that we'll be putting out.

[Host 2]

What about on the other side? I've heard there's a, like, a really way to. Really great way to make a lot of money really quickly as like an unlicensed broker dealer of like, SPVs. Like, would you recommend that? What's your take? You know, get into that game, like, find. They go to some employee.

[Host 1]

There's no securities laws.

[Host 2]

Random person. Oh, there we go.

[Delian Aspro]

Can't find me there.

[Host 2]

But how have you been processing the spv? Boom. All of these questions around brokering and what's going on there.

[Delian Aspro]

Yeah, I mean, I think, I do think that one lady's like anthropic tweet basically led to anthropic's like, policy on this. Part of why that sort of is happening is there's this SEC regulation on the number of entities you can have on your cap table while staying a private company. I forget the exact number. I think it used to be a thousand, now it's 6,000. These private companies, as they stay private so long and have this number of SPVs, and especially as these SPVs get warehoused out, it increases. Increases your entity count. And so I do think there's a significant, you sort of lockdown on it.

[Host 2]

And employees typically own options and they don't show up on the cap table until they exercise.

[Delian Aspro]

Exactly, exactly. So, yeah, I mean, we're obviously tracking that of Art in terms of how many entities we have. We're not seeing some like, exponential growth. We're like, we're too worried about it. You know, I don't know that we've been like, as stringent to date, but as I look at our future rounds. Yeah, like, you need to start to think about, like, okay, how many more rounds am I going to stay private? And at some point, point, like, I need to keep the entity count, manage and say, I totally get the anthropics of the world. It's like they've just raised so many good bajillion of dollars in the private markets that like, they have to be.

[Host 1]

I did get some more backstory on that one. One deal. Yeah, it was allegedly. Allegedly. It was a much more like complicated structure that wasn't as like bad sounding as the original post said. And like there was like an SPV involved and like some other stuff. But anyways, why? Every single person that you know that has invested in the company has probably come to you saying like, what are you doing with data centers? And all the biggest companies in the world are now like dedicating resources to it. You guys are seemingly staying like extremely focused. You've had conviction around timelines around space data centers being longer, but now that the sort of regulatory environment on earth is changing, things that could change potentially the need or the economics. When would you change your mind and capitulate?

[Delian Aspro]

I mean, I'm definitely open to changing my mind. As a shareholder of companies that may take advantage of these technologies, as somebody that's going and building a space company, I really want to focus on like where I think we have the right to win and can sort of build up something that has like, you know, sort of durable margins.

[Host 1]

Yeah, not, not a great situation. Like you being in a situation where you're like this like deeply entrenched partner with five, you know, massive pharmaceutical companies that are dependent on you for a specific part of their supply chain. That is much better business to me than like competing with SpaceX on like, you know, putting GPUs in space when they just are naturally going to have a lot of advantages. And so it feels like, yeah, you could pull forward like around. But do you shoot yourself in the foot in the process of that?

[Delian Aspro]

Yeah, I just think there are a lot of companies in the space industry that will raise off of contracts, renders, etc. But ultimately these things aren't durable unless you are regularly flying spacecraft building up a production line. Right. You know, the Elon quote that I love is like the, the factory is the sort of product. I think the thing that we've done well at Varda is we've stayed hyper focused on just like our particular W series vehicle, getting that up to production rate, flying it regularly, making sure that it's succeeding and serving customers across a wide variety of verticals versus distracting ourselves with like, there have been a lot of opportunities that we've said no to. I'm sure some of those things in the short term would have inflated our valuation, top line, et cetera.

[Host 1]

But wearing your investor hat, you're like, I don't care about what my valuation is today. I care about, I care about where

[Delian Aspro]

it is in like 10, 20 years. And so like, I don't need to like listen to the noise of the market. And so I think that is the benefit of like having somebody like me on the sort of cap table or as a, you know, sort of co founder versus like there are plenty of companies that had XYZ vision two years ago and are now fully pivoting to space data centers because like, that's where investors are deploying capital. And at some point it's like if you're changing your company vision and product off of the basis of like where investors are like cycling capital into versus like what customers are demanding. And this is where it's like, look, are there areas where it's like, yeah, Zblan might get affected? And there I think we have much more of like a right to win. And that affects terrestrial data centers and things like that. Yeah, like there might be areas that we sort of play but like go launch a bunch of Nvidia Jetsons up into orbit and do that at economics that aren't competitive relative to SpaceX. And even then I think the SpaceX economics, while promising, still have a long ways to go. That's not a game that I want to play in. You know, competition is for losers.

[Host 1]

There you go.

[Host 2]

Yeah, yeah. I mean also, you're so far from realizing the TAM of what you're building. It makes sense for Google to jump into AI. Like search is saturated and they need the next thing.

[Host 1]

Yeah, it's possible we'll look back on this moment and be like, okay, every viable competitor to actually go and compete in pharma went and did this other thing. And maybe some did well, but then that still left an opportunity to just kind of like verticalize, own the category.

[Delian Aspro]

And I think what like tech people misunderstand is just like the like size of the TAM in pharma. Oh, yeah. So, you know, anthropic is what I think like a $60 billion run rate or something like that. If we get like the, you know, sort of latest numbers. Keytruda, that drug that I mentioned that Merck did in 2019, that one drugs, top line last year for Merck, $25 billion in revenue. Just that one drug.

[Host 1]

Yeah.

[Delian Aspro]

They obviously have many other drugs. There are many other cancer drugs, there are many other immunology drugs. And so it is a massive tam. If you're able to actually significantly improve the performance of these things, capture some upside, start to go more full stack yourself in pharma, I'm not concerned. And especially by the way, once there's AGI, what do you think everybody's going to be focused on Afterwards when like there's no more like, you know, white collar, you know, sort of, you know, labor that is needed, everybody's just going to want to live for forever. Where do you think all that capital is going to rotate into? It's going to rotate into longevity, therapeutics, manufacturing. And so I'm, you know, sort of happy to stay on the sidelines and sort of accept that capital once all the AGI people rotate out.

[Host 2]

I love it. I love it. Well, thank you so much for coming on the show.

[Delian Aspro]

Oh yeah. Excited to finally come in person. Good to see you boys.

[Host 2]

So good to have you in person. We'll talk to you soon, darling. Have a great.

[Host 1]

Say hi to the team.

[Host 2]

We have Richard Socher from Recursive in the waiting room. We've been keeping him waiting so we will bring him in to the TVPN Ultra Dome. I love when companies emerge from St. Delft. Nothing makes me more frustrated than one day.

[Host 1]

What's going on?

[Host 2]

We've talked to Richard before, so yes, welcome back. But we're very excited.

[Host 1]

We've talked to you before.

[Host 2]

Yes.

[Host 1]

That was a bad joke. You dot com, right?

[Host 2]

Oh, that's right. That's right. Yes.

[Host 1]

Great to see you, Richard.

[Host 2]

But even though you've been on the show before, let's reintroduce the company and sort of set the table for us on what you're building.

[Richard Socher]

Happy to.

[Host 1]

Yeah.

[Richard Socher]

So still happily running U.com and Aix Ventures but now very excited to start recursive with seven incredible co founders. At Recursive we're looking to build Recursive self improving superintelligence to automate knowledge discovery. It's a lot to unpack there. You know, we're getting closer and closer to AGI. I think we have to set our sights, set our goals now to superintelligence. I think AI is code and I can code. So now we can actually put those two together and allow in an open ended fashion to let AI actually improve itself.

[Host 2]

Yeah. Are there any other, is the entire focus on AI research or are there any other areas of research that you're interested in? You see a lot of labs do research on cybersecurity or research on bio or science. There's a whole bunch of different initiatives. How focused do you want to be on AI research?

[Host 1]

Yeah, it is part of that, part of that focus. Just if you can get RSI actually going, then everything else you can potentially seemingly very easily.

[Host 2]

There you go.

[Richard Socher]

You kind of answered the question. That's exactly right. I think for now as a company you want to have a very strong laser focus. And that laser focus for us is on enabling AI to automate the scientific method on itself. The ideation, implementation, validation of ideas.

[Brandon Hill]

Right.

[Richard Socher]

And once it's gotten really, really good at that, that means it can be fully recursive, self improving. Then we're going to apply it to a whole host of other applications, first in software and broadly construed digital work, eventually even physical work. And I saw your previous show just now. I'm very excited about preclinical biology and the impact AI can have on that.

[Host 2]

Talk about the round, the structure. Who's in? How much did you raise?

[Richard Socher]

We raised north of 650 million.

[Host 1]

Thank you.

[Host 2]

There we go.

[Host 1]

Walk us through the pitch that you gave investors. What's your right to win? You're one of many teams working on this problem, including all the existing labs. But I'm sure that the pitch was compelling, otherwise you wouldn't have gotten a $650 million.

[Richard Socher]

That's right, yeah. At a 4 billion pre, as led by Google Ventures, Greycroft, some amazing participation, also from Nvidia, AMD and many other incredible funds that we're very grateful for, Angels, strategics and so on. Our right to win is essentially fourfold. It's the focus. Yes, we know that everyone works a little bit on self improvement, auto research, things like that, but no one is fully focused on this. We have the team, we have incredible co founders, Tim Rochteschel who invented retrieval, augmented generation with others and Rainbow teaming and prompt reader and Genie 3, which is like the most sophisticated world model that Google DeepMind had recently launched, like some incredible talent. We got Jeff Klune who has worked together with Tim R many things he's done incredible papers like the Darwin Google Machine, Hyper Agents with some of his students who are also with us. We got Tim she, who built Cresta into Unicorn AI for service centers. We got Josh Tobin, who's one of the earliest people at OpenAI and eventually led some of their work on ChatGPT agents, Codex Deep Research. We got Saming Xiong, co inventor of prompt engineering. We got Alexei Dosovitsky, co inventor of the Vision Transformer. We got Yuan Dong Chen who used to lead RL at Meta. I mean, it's just such an incredible team.

[Host 1]

Yeah, that's pretty compelling.

[Host 2]

Then how are you thinking about scale being all you need or not? Is there a capital war? Or is $650 million enough to do the level of training or inference that you need to actually reach rsi? Because if a lot of the labs at least are sort of putting the AGI ASI terms in orders of magnitude of compute or orders of magnitude of dollars. And so is there some sort of race condition or do you think that there's a more like elegant solution? We need new algorithms, I guess it's a great question.

[Richard Socher]

Certainly 650 in this league that we're now in is step one. If you want to compete at that level, you want to build massive frontier AI, you have to eventually raise more. At the same time, this will last us for quite some time unless we're seeing enough progress that makes us want to accelerate this. Right? And then we prove a new kind of scaling law where more compute results and more inventions, more capabilities in AI, more automation in the process of AI research itself, then we're probably going to accelerate that. And to be honest, the team has already made some incredible meta inventions that led the AI to make its own inventions. And so we're really excited about the progress and likely will accelerate.

[Host 1]

Yeah, do, do you view today's coding agents as somewhat of a form of rsi or is that the wrong, you know, kind of wrong category?

[Richard Socher]

I'm glad you asked that question. The current models are great for auto research, which is a step towards rsi, but you're not not doing true recursive self improvement if you don't have full control over the entire self, namely the model, the architecture, the infrastructure, its entire harness and all of the things that go into the training pre mid post training. All of these things. And so they are. It's a useful step to use coding agents for auto research, but it's just a step towards the true RSI vision.

[Host 1]

Got it.

[Host 2]

How are you thinking about standing on the shoulders of giants broadly? I mean you mentioned a bunch of researchers who have published a bunch of papers and obviously you're integrating that. But is there a way to pull forward progress or catch up to the frontier faster based on open source models, open source research, open source data sets, or distilling or integrating existing open source projects? I would love to know just your thesis on leveraging open source and tweaking and fine tuning and iterating on versus start from a completely blank sheet?

[Richard Socher]

Great question. Yeah, I think as with many things and similar to the applications of rsi, you want to start with a laser focus and then open up the aperture to more and more things. And 100% you want to stand on the shoulders of giants, both like closed open source. Open source has done an incredible job bringing in more capabilities to the community and so we will, we'll use anything and everything that we can. But we also know that long term you're going to want to own and be knowledgeable and able to build the entire stack.

[Host 2]

Yeah.

[Host 1]

Now that you have your own funded neolab, do we need more new NEO labs or do we have enough?

[Richard Socher]

I'd like to think so. One, we want to build a real company.

[Host 1]

Right.

[Richard Socher]

We're not just a lab. I sometimes struggle a little bit with that terminology. I think there are a couple of labs that are truly just labs and it's a little bit unclear where they're going to go long term. We have folks that also build unicorn companies from scratch with revenue and so on, have launched amazing products like an OpenAI Codex with Josh and others. So we're excited, but at the same time, this machine we're building is almost like a Eureka machine in the sense of like, it'll keep making inventions for you. And so a lot of labs that are focused on like, continual learning and world models and maybe memory and all of these other aspects. I believe those are all important aspects. I believe that if we're successful, those are all just special cases and outputs of this machinery that we're building.

[Host 2]

So do you. Yeah. I mean, how close are you? I imagine that this is a whole process, but how close are you to thinking about a product for businesses or a product for consumers? An API, a web playground, a demo? Is that something that by design you want to be super fast about and get something out into the world, have some impact in the interim, or do you want to keep everything closely held for potentially a really long time? I don't know, forever, potentially? If you're just going to invest.

[Host 1]

Yeah. When I hear the Eureka machine, I'm keeping that to myself personally.

[Richard Socher]

Initially, we thought about really giving ourselves a year or more of time. In the last few months, the team has already made some incredible breakthroughs. And so we may accelerate some of the productization of this research to come out earlier. I don't want to commit yet to specific date or anything, but, like, I think, you know, we want to build a viable company here that brings super intelligence and allows it to have massive impact to help humanity flourish. And that won't happen in just the research lab.

[Host 2]

I love it.

[Host 1]

Could you ever see a scenario where you create something that would make more sense to, like, spin off into another company to productize? Or do you. Do you expect to basically use all of your creations internally and productize them yourself?

[Richard Socher]

I can see how our customers will use this in incredible Ways so that they can then use like create new kinds of products and new kinds of product categories. I think we've seen similarly incredible new companies like that that have used AI and code generation to enable others to build their own businesses. You know, replied lovable and so on are great examples of that. And I can see our technology doing something similar at an even more foundational level.

[Host 2]

Amazing. Thank you so much for coming on the show. Congratulations on the product.

[Host 1]

Great to get the update and come back on for your first release. Yeah, great to see you.

[Host 2]

We'll talk to you soon.

[Host 1]

Cheers, Richard.

[Host 2]

Have a good one. Goodbye. We have a debate. We have a debate.

[Host 1]

What's that?

[Host 2]

In the Wall Street Journal, they are reporting that typing is being replaced by whispering. And it's way more annoying. This is a trend because programmers, startup founders, startup employees are starting to resemble high end call centers. Only these employees are talking to AI and people are saying it's a little awkward. Where do you stand on whispering? Tyler, you have a take.

[Tyler]

Well, so I think this might be a play on words, right? Because whisper is like the model you use to speak into. Yes, but apparently whispering.

[Host 2]

They are literally whispering. So Molly here mumbling. Molly's mumbling was starting to get on her husband's nerves. What was once a sacred nightly routine, putting the toddler to bed, collapsing onto the couch and opening their laptops to finish their work in peace, had become anything but peaceful. Instead of typing quietly, Molly started to hold down the function key and talk in hushed tones to her computer. She runs her own AI business in Seattle and is hooked on Whisper Flow company. You might have heard of a dictation app that users are pairing with coding tools like codex from OpenAI and Claude Code to turn rambling stream of consciousness prompts into coherent, usable texts in seconds. Efficient, the Wall Street Journal says. Yes, but annoying? You bet. It didn't take long before Molly's husband told her they needed to talk. The couple now often sit apart. If we need to get something done at night, one of us will stay in our office. Across Silicon Valley, work is being remade as once mellow spaces become dens of din. And we got a mention of our friends over at Ramp. One venture capitalist said visiting AI startups today is like showing up at a high end call center. Everyone is except everyone is chatting with AI Engineers at credit card startup Ramp sit at their desks wearing gaming headsets so they can talk loudly to their AI assistants. Gusto co founder Edward Kim has encouraged employees at the human resources company to experiment with dictation. Technology telling them the office of the future will sound more like a sales floor. People are going to be screaming at the AI agents.

[Host 1]

Oh, I like that. Like, yeah, stock trading.

[Host 2]

Yeah. We're bringing back the New York Stock Exchange. There is a way to combat this. I've seen a couple masks. The Bayne mask, the Bain mask. It's the Bane mask. Basically, people do put it over, and it has a headphone jack and a USB C port, and then they can yell very aggressively if they're live streaming video games late at night and they don't want to wake up their partner or roommate. But I think I'm pro this. What do you think, Jordi? I think I'm in favor of a noisier. Louder. I do like a library environment. I like a calm environment, but there's something invigorating about everyone yelling at their computers all day long. Where do you stand on it, Jordy?

[Host 1]

I like it.

[Host 2]

You like it?

[Host 1]

I bring in the energy.

[Host 2]

Yeah. What about you, Tyler?

[Tyler]

Yeah, I like the aesthetics of people yelling. Although the muzzle coating. Muzzle, I think, is a good.

[Host 2]

You won't put on the Bane mask.

[Tyler]

I don't want the muzzle.

[Host 1]

No.

[Host 2]

Yeah. You don't want the muzzle. Yeah. You can't muzzle me.

[Tyler]

You're not going to stop me from screaming at my computer.

[Host 2]

So you say you're a fan of this, but I don't feel like I've ever seen you talking to Whisper Flow while you. You are vibe coding.

[Tyler]

That is true reveal preference. I find I usually only use the speaking when I'm, like, driving or something.

[Host 2]

I do that all the time. I'm always on the dictation mode. I don't use the advanced voice mode that often. What I will do is I will go and press the little microphone button, ramble for a few minutes, and then fire off a more robust query to ChatGPT with more context.

[Host 1]

Yeah, you're doing that, like, once a day, and it's incredibly powerful. I. I actually don't think many people realize that you can just ramble once

[Host 2]

a day in front of you, like 50 times a day.

[Host 1]

There's some controversy. Has hit the timeline.

[Host 2]

Okay, what's going on?

[Host 1]

Socks has pulled a screenshot of the figure livestream where the robot appears to be doing something like adjusting a headset.

[Host 2]

Okay.

[Host 1]

And so he's alleging.

[Host 2]

Oh, he thinks it's teleoperation.

[Host 1]

It's teleoperation. And the robot is actually just being controlled by someone in a VR headset. There are some comments that say, hey, maybe this could have Been in the training data.

[Host 2]

I was about to say. Yeah, it could be in the training data. Interesting.

[Host 1]

So who knows?

[Host 2]

It's interesting to think about all the vestigial behaviors of human behavior that might exist in training data for humanoid robots. It might just go and pick up a phantom Diet Coke and take a sip because it's had a long day and it knows that's what humans do. So it recreates that. We'll let Brad Adcock respond to this. Dunk on it.

[Host 1]

Yeah. The main thing is the speed. The speed up is insane. Sorry.

[Host 2]

I don't know.

[Host 1]

Just the robot, how quickly it's moved.

[Host 2]

Yeah, yeah. Because the last version, he was talking about the Biden walk going away, and it was still a little jerky. And this doesn't look jerky at all. This looks very polished. And I don't think we've seen anything at this level, like Optimus is still a little jerky. And even the unitary stuff is pretty jerky. That crazy huge Unitree was pretty jerky. And we're all sort of used to that because it doesn't really matter if

[Host 1]

jerky is getting the job done. I do think they should have allowed some type of actual live demo so people could go and see the robot

[Host 2]

saying, hey, or like a respected journalist in the room or something. I don't know. There's a bunch of ways. But this. This debate will continue going back and forth with people understanding and debating the capabilities of this technology. But we can come back to that in the future because our next guest is in the waiting room. We have Brandon Hill from Vori, the CEO. I kind of like that. How you doing, Brandon?

[Host 1]

What's going on?

[Host 2]

Welcome to the show, guys.

[Brandon Hill]

Thanks for having me.

[Host 2]

We gotta ask you, what's your stance on whispering in the office? Have you heard this? All the AI employees, all the startup employees are whispering into their computers to give commands to Claude code and codecs. Are you in favor of having your office feel like a sales floor, or would you like a quiet library, like, environment?

[Brandon Hill]

I mean, ABC always be closing, right? Like, whether you're closing PRs, whether you're closing deals. We want Kinesis on the.

[Host 2]

I love it. I love it. I love it. Yeah, I agree with you. I think everyone, it's unanimous.

[Host 1]

Get used to it.

[Host 2]

But we're not here to talk about whispering your computer or screaming at it. We're here to talk about your company. So please give us an introduction on yourself and Vori.

[Brandon Hill]

Yeah, absolutely. Guys, thanks for having me. My name is Brandon Hill. I am The CEO and co founder of Vori. I'm actually a third generation grocer. Just to give you my story.

[Host 2]

That's amazing. Yeah.

[Brandon Hill]

My grandparents were in grocery.

[Host 2]

Cool.

[Brandon Hill]

My parents met and fell in love in a grocery store and have been in the industry for over 40 years.

[Host 2]

That's me.

[Tyler]

Amazing.

[Brandon Hill]

I went to Stanford University is where I met my two co founders who are two of my best friends. Worked at Google with Trey. Rob was a aerospace engineer at SpaceX under Elon and then self driving cars at Lyft. And we're building the operating system, a self driving operating system for supermarkets.

[Host 2]

Okay.

[Brandon Hill]

Where they basically help them make more money.

[Host 2]

Yeah.

[Brandon Hill]

And operate more autonomously. Because as you know, grocery is a paper and pencil industry, but it's also $1.5 trillion of the American economy.

[Host 2]

Yeah. So are you starting with back office erp, inventory management, accounting, or are you going to try and replace those automated checkout machines that everyone knows barely work even though they've been around for decades? It feels like there's opportunity in both. How quickly do you want to focus?

[Brandon Hill]

Well, you know what's interesting is we found that the alpha is really in automating everything before checkout.

[Host 2]

Okay.

[Brandon Hill]

And after checkout.

[Host 2]

I can imagine, as you know, there's

[Brandon Hill]

a bunch of companies that raised hundreds of millions of dollars trying to do the just walk out Amazon model.

[Host 2]

Yeah.

[Brandon Hill]

The unintuitive thing is what you're surprised by is it actually turns out building inventory, automating pricing.

[Host 2]

Sure.

[Brandon Hill]

Helping grocery stores market their products to their customers. Everything before and after is really where the, that's where the economics are and that's where the grocery stores have the most need and willing to spend money. So that's what we do. We happen to have a sleek, modern checkout on the front end, but it's about connecting everything in the store end to end.

[Host 2]

Okay. Automating pricing in a store. There's the price of oil's going up, there's inflation, there's all sorts of different prices that are changing all day long. The price of milk fluctuates day to day. You mentioned a pencil and page, paper and pencil industry. I imagine that at some point a store manager gets a spreadsheet or an email or some sort of notification that a price needs to be changed and then they have to print out a new label. Is that how it works? And then how will it work in the future?

[Brandon Hill]

So there's the old way and the new way.

[Host 2]

Yeah.

[Brandon Hill]

In the old world, imagine a supermarket which has 100,000 products whereas a grocery store, you know, a restaurant might have 50100 menu items. Grocery store has 50, 100,000 SKUs. They're accepting dozens of payment types.

[Nate Tepper]

Right.

[Brandon Hill]

They have hundreds of employees and they might do more in sales than a restaurant or like a donut shop or a coffee shop might do in a month or a year. Yeah, in terms of overall volume. So the complexity of solving the problem is really hard. Now today, what they're the state of the art, if you could believe it, is you have a pricing coordinator like named Judy. She has her bifocals on. She receives an invoice from the supplier. She's looking at it with her bare eyeballs, going through line by line with a ruler and highlighter to figure out if milk and bok choy and ketchup and cereal have gotten more expensive today versus last week. She finds the price change and has to go type it into her ERP system like the back office system and hope that that reflects accurately at the POS. It's 12 step process. Vori does it in one click. So our pricing agent reads all the incoming invoices from all of the hundreds of wholesale suppliers that a supermarket is using. It detects when there's changes in cost, which of course there always are due to inflation and supply chain issues, and then automatically changes the price across 3000-0500-0010-0000 items in real time.

[Host 2]

So the changing of the price at the checkout actually seems like the easy part because when I'm in a grocery store, I feel like a lot of the prices are printed and they're labeled right on the shelves. Is there a movement towards E ink displays or LCD displays that could be updated in the cloud in the future? Is that an important step along the road to automating the price setting of grocery stores?

[Brandon Hill]

Oh, it's critical. In fact I got a couple right here. Electronic shelf labels, right? Walmart actually just spent a billion dollars rolling out electronic shelf labels across every single item across every single of their 4,600 stores. But the thesis behind Vori is we do that today for all of our customers and we're taking like that Walmart and Amazon level technology and rolling it down market to other 75% of the trillion and a half dollar food and bever retail market.

[Host 2]

I didn't realize it was that fractured. I would have guessed do grocery stores doing like 70% of all grocery or something?

[Host 1]

Do some of these smaller grocery stores or I guess the full stack end up selling products at a loss accidentally sometimes because they get a price Change.

[Host 2]

And then it didn't get rolled out fast.

[Host 1]

Yeah, they didn't roll it out. And then they're just like losing money on an item. Is that, is that a thing?

[Brandon Hill]

Well, it happens all the time. So Dylan, right? Dylan from tvpn. So Dylan, I was going back and forth with him on X yesterday. And so his favorite shop is Mill Valley Market. Okay. Up in Mill Valley. Love that store.

[Host 2]

It's beautiful.

[Brandon Hill]

That's powered by Lori before we went in that store and Ryan, the owner will tell you, yes, they were mispricing items here and there and they were losing margin. We go in, we find 5 points of gross margin sitting on the table with our. And give it back to them. Put it back in their pockets with our automated.

[Host 2]

Take it out of Dylan's pocket and put it in the store owner's pocket. I like this.

[Brandon Hill]

Hey, I'm pro competition. I'm pro America. That means I'm pro small business.

[Host 2]

Right?

[Brandon Hill]

Just trying to help them succeed.

[Host 2]

No, no, it was doing fine. But yeah, it's funny to frame it that way.

[Host 1]

The price of bok choy is going up.

[Host 2]

Yeah. Good luck. Dylan. Get ready to get ready to pay through the nose for your bok choy. Tell us about the round. How much did you raise? What happened?

[Brandon Hill]

Yes. So we raised $22 million to.

[Host 2]

Thank you.

[Brandon Hill]

We raised 22 million to invest really in our go to market. We been expanding rapidly. We've done over $500 million in payments.

[Host 2]

Wow.

[Brandon Hill]

Which is where 60% of our revenue comes from. We're investing behind that, advancing our product, releasing more agents for the store to do more work to help them be more competitive against big, big box stores. And investing behind our deployment team to help install this hardware and software and complex payment system into more grocery stores across the country. So we're super excited.

[Host 2]

That's amazing. Well, congrats. Seems really smart and very, very unique market to. Very unique.

[Host 1]

Hard to compete with you, having probably done all of these different jobs at different points yourself back in the day.

[Host 2]

Generational success. I love to see it. Have a great rest of your day.

[Host 1]

Great to meet Brandon.

[Host 2]

Thank you so much for coming back on soon. Yeah, we'll talk to you soon. Have a good one. Up next, we have Nate Tepper from True Drew Short. He's the founder and CEO Here to talk about funding from Coastal Adventures Wonder Co. Bunch of Hollywood investors for AI generated vertical films for mobile. The Tepinator. Welcome to the show.

[Host 1]

What's going on, dude?

[Delian Aspro]

How are you doing?

[Nate Tepper]

Hey, guys. Good to see you. Let's go.

[Host 1]

Let's go.

[Host 2]

First time on the show. Introduce yourself, Tell us about the company, Tell us about the product.

[Nate Tepper]

Nate Tepper, founder, CEO of True Short where a film studio and streaming platform. So on one side we have creatives and engineers that are making movies together. On the other side, we make the streaming platform too, like Netflix. Distribute the movies, monetize them. Full stack creative company, vertically integrated.

[Host 2]

What is the. Like, how do customers experience this? A mobile first, web desktop. Like, what is the watching behavior these days?

[Nate Tepper]

Yeah, exactly. So we were kind of inspired by the micro dramas. I don't know if any of you guys know about that, but basically saw that China was pumping these vertical soap operas that were making more money than the box office in China. Like some of these apps are doing 50 million a month in revenue and they're all shot vertically, minute long episodes. And that kind of led us to this idea that like, wait, there's something in between Netflix and TikTok. People are consuming this on their phone vertically, short form. So we kind of took that exact same approach, but not.

[Host 1]

Yeah, we had that realization probably a year ago too. We were looking at some of the iPhone kind of charts and seeing like, what are all these companies I've never heard of that are in the top 25 that are, you know, printing money? Yeah, is this. And someone else had a realization like this years ago. One of your investors, good friend of ours, Jeffrey Katzenberg, you got him into the round. What did he see in your approach that would make this a great investment?

[Nate Tepper]

I mean, the next Hollywood is being built right now from the ground up. Like all these new creators are coming in. I think some kid in his bedroom is going to make the next Star wars, leveraging AI and maybe different formats, short form, etc. And I think he just sees that opportunity and the approach that we took. We have this vision of like the next Netflix, but we launched with a true crime app. We just took a very strategic, like Amazon launch with books. We took a niche wedge approach and within six months with the number one true crime app, 3 million in annualized revenue, I think he saw the vision, but that actually we know how to execute as well. And so I think he was bought into that.

[Host 1]

How are you working with creative talent? Are you hiring all over the world? Or is this going to be like, you want to get a big studio here in LA and hire a bunch of people?

[Host 2]

The next James Cameron?

[Nate Tepper]

Yeah, let's go. I like that. We actually just opened an office in la and so we call them creative pods. Is kind of the way we work. A pod is basically a showrunner who's like the visionary, knows the story, the characters, the brand. Then there's an AI filmmaker who brings all the shots to life with AI, kind of like the DP director of photography. And then you have the editor who strings it all together. And that pod basically produces content every week. A series a week. And so we want to keep building more and more of these, these creative pods. And we're testing a bit right now. We have remote pods, but we just opened our first office in LA with our in person pods and just the speed that you can work collaboratively together there. So we're stoked about that.

[Host 2]

Talk about the customer acquisition, Flywheel. You've been at the top of the app charts. I'm sure that that has a lot of energy. There's news about the company, but how do you think about the content being shareable? One of the critiques of Quibi was that you couldn't screenshot or, or screen record the content at the same time. If someone's just wiring up every time something goes out on True short, just upload the full thing to YouTube and TikTok and Instagram and like, let me steal the ad revenue. That's not going to work. So how are you, how are you dealing with getting attention for the company? Having the Flywheel, leveraging the content that you're creating without giving everything away and still giving customers a reason to subscribe and pay you?

[Nate Tepper]

Yeah, yeah, Great question. And like, growth is in our DNA. Jordy knows this, but our last app was number one in the App Store for health and fitness. I think it had close to a billion views on TikTok. Let's go. And we, I mean, look, we run like kind of similar playbooks. We're super growth hacky. We're running many different TikTok accounts. Like, we know how to do all that stuff.

[Host 2]

Okay, Steal it yourself. You steal content yourself. That's probably the right move. Yeah, I mean, that's what we do here. We clip our own show.

[Nate Tepper]

Yeah, I know. We make content. We might as well put it out there everywhere. Yeah, but, but I do want to be straight up to like, the micro drama space is run on paid acquisition right now. Very much so. And like, I won't deny that too. We, yeah, we definitely run paid acquisition. We just know how to do both.

[Host 2]

Yeah.

[Nate Tepper]

So we have a growth advantage.

[Host 2]

Yeah. And I feel like the, the cost of paid acquisition probably falls when there's already some organic content out there. There's more familiarity with the brand. All of that helps with the CPMs.

[Host 1]

Yeah. How quickly do you expect the YouTubes of the world, the TikToks, to actually roll out dedicated micro drama features?

[Nate Tepper]

TikTok did. Okay, so TikTok has one called like Pine Dream Pine something. I don't know, I don't know why. It has like 200 ratings on the app Store, so I don't know what

[Host 2]

they did with it.

[Nate Tepper]

Netflix is creating kind of like vertical section in the app, but they haven't made vertical content and yet. So it's definitely happening. Things are going.

[Host 1]

Yeah, feels like. It feels like this business to me will over time probably look more like a 24 that owns its own distribution and actually sells the. The product. Because you're going to get like competition from all the platforms, all the features. And so your advantage is like, can you create a system that can reliably create hits and IP and deliver enough value to customers so that they subscribe?

[Host 2]

They keep coming because YouTube. YouTube had.

[Host 1]

Is that, is that, is that, is that right?

[Host 2]

Yeah, yeah.

[Nate Tepper]

That's why we want to build the full stack here. Like it's almost like its own flywheel where we create the content, put it in the app, funnel that data back into our own technology and creative teams and just keep making more. So the speed that we're going and the learning, the pace of learning on the creative side too, I think we'll get to hits much faster. I don't honestly just don't see these companies keeping up in terms of speed and iteration. Not of creative, actually.

[Host 2]

Yeah, yeah, Yeah. I mean, YouTube got into original content that they were producing and they pulled back from that eventually and they just maintained just pure platform. Netflix obviously does productions and. But a lot of that is done with other studios and then they just buy or license the, the actual end product. So, yeah, the vertically integrated approach.

[Host 1]

Yeah, I think, I think the key will be creating a setup where a pod can make more money, like building on True Short than they can by just going and setting up their own kind of YouTube account.

[Nate Tepper]

I'm. I'm curious in general, like, part of what I'm thinking about is like there's about to be a million filmmakers. Like some, like I said, some kid in his bedroom is going to make the next Star Wars. Where is that going to be distributed? Is it YouTube? Is it Netflix? Is it True Short? I think we're going to see emerging films, which is already happening, that are actually high quality, quality that people want to watch. Where are people going to Watch them? I don't know yet. I think we'll see some stuff over there as well.

[Host 2]

How are you thinking about episode length and number of episodes? What have you actually tested? It feels like the base case is like a two minute episode, maybe ten episodes per story. Is that roughly correct? And then do you think there'll be wide variation? Like we see a 90 minute movie and then a three hour epic from Nolan every once in a while? Like where do you see the dials turning?

[Nate Tepper]

Yeah, I mean I actually view this microdrama space and what we're doing as like separate from theatrical, even separate from Netflix. The way people consume it is different too. It's like they open this stuff when they're bored on the subway. Right. And so it's actually, if you still want to go watch a full two hour length movie, you're going to go watch that on Netflix or go to the theater that will still exist until these become good enough that they're distributed there. But, and so for now, just for now that we're focused on these like kind of niche genre, one or two minute long episodes, many episodes per series and like getting a flywheel going around that.

[Host 1]

What is, what is the. Do you feel like you have hit IP yet, something that you'll invest, you know, $20 million into a specific like series over time yet, or are you still just like taking, taking shots, getting base hits, doubles, etc.

[Nate Tepper]

So I do intuitively we're launching it later this month. It's not true crime, so it's not on our True Crime app. It's our next genre. And just seeing what we're creating there, I would honestly say our true crime stuff is just like that will be the lowest quality we'll ever put out. Stuff like it's getting better and better. So intuitively I think we're making now is real IP people that will have fans.

[Host 2]

So True Crime, True Short, there's obviously a linkage there, but the name True Short feels like it can hold anything. Like it's not like home box office. No one cares that. They just think hbo. And it feels like you'll be there very quickly, but it sounds like you're considering multiple apps. Is that in the roadmap?

[Nate Tepper]

I don't want to say too much yet about exactly what direction because we're still discussing it. But what I will say is there are a lot of opportunities now around niche genres.

[Host 2]

Yeah, I think that's interesting. What about pseudo or semi user generated content? Hbo? Yeah, you need an agent. It's incredibly High bar to get your media product, your movie or show on hbo. Amazon prime sort of has a self publishing route. If you're really creative and Kindle self publishing, you can publish a book without any. I think it needs to meet standards, but it doesn't necessarily. It's like an app store review. It's maybe just one notch above YouTube to publish on the Amazon suite of products. How are you thinking about engaging like that random. The next James Cameron who isn't in LA and didn't have a great resume, but created a great product and you could monetize it for them. Have you thought about that?

[Nate Tepper]

Yeah. I mean, you nailed it. That was the original vision, was like, let's be the distribution for these million filmmakers that are coming. And curation and taste is going to be really important things like that. But we quickly learned like, and just realized that like out of those million filmmakers, a lot of it will be slop. And it's still the earliest days. And we also learned from the micro dramas that like there's a formula especially for winning in the beginning. So we're like, let's own it, let's learn it ourselves, let's build it right now. And we can take this. We could probably take this to a billion dollar company just as our own team moving fast. But I do think the big, big vision, like I said, is like, how do you find the next James Cameron? Where is he putting his stuff? That's. That's really exciting.

[Host 2]

Favorite song.

[Host 1]

Well, we have 30 seconds left. Do you want to use the opportunity to do a quick solo or. Or your guitar?

[Nate Tepper]

I knew you were gonna ask me to do that. But.

[Host 2]

But I will plug.

[Nate Tepper]

I will plug. Check out my band Pink Roses on Spotify. If you know Dave Fontenot's my bandmates.

[Host 2]

Oh, amazing. Yeah, we had J on the show.

[Host 1]

Yeah, they've had this crazy side quest of being rock stars.

[Host 2]

Amazing.

[Host 1]

It's amazing actually.

[Nate Tepper]

It actually, it has millions of streams. You like it?

[Host 2]

Well, congratulations on the round. I don't know if we hit the gong.

[Host 1]

Hit it again.

[Host 2]

Boom.

[Richard Socher]

Thanks, guys.

[Host 1]

Great to see you, Nate.

[Host 2]

We'll talk to you soon.

[Host 1]

Make a micro. Make a micro drama about Tyler the intern's life at tvpn Me.

[Nate Tepper]

Should we use his likeness?

[Host 1]

Yeah, yeah.

[Nate Tepper]

This is recorded.

[Host 1]

Yeah, there's a lot of training data too. Sorry about it.

[Host 2]

Well, later.

[Host 1]

See you, Nate.

[Host 2]

Thank you so much. Our next guest is the co founder and CEO of Run Roadrunner, a new company for AI native revenue infrastructure.

[Host 1]

I like the sound of that.

[Host 2]

John, welcome to the show. How you doing? Hey guys, thanks so much for joining. How you doing?

[Host 1]

It's happening. Good revenue. Infrastructure. Revenue. And I love infrastructure.

[Host 2]

Yeah, we talked a couple years ago. But for those who don't know you, please introduce yourself in the company.

[Jubin Murzadogan]

I'm Jubin. You guys pronounced my name, right? Both first and last, which I was impressed by. I am one of the co founders and the CEO of Roadrunner. Thanks for having me.

[Host 2]

Yeah, nice to meet you. I mean introduce the product, introduce the problem and sort of explain where you fit into the businesses that you're selling into these days.

[Jubin Murzadogan]

Sure. So the backstory is that we incubated the company at Kleiner Perkins. The reason that we did it was because I run a group of CIOs here that meet twice a year and over the course of several dinners they basically told me it's the most broken problem inside their company. And I actually didn't believe them because I was like, there's no way this has not been a solved problem. And they were like, no, seriously. And in some cases it was the number one negative NPS surveyed software inside of their business. I come from sales, so I've been the consumer of this problem. And I'm like, okay, I understand the problem. So we did a full market map, looked at every company, ended up not finding anything, and so decided to do it ourselves.

[Host 2]

So you said that the existing solutions were low nps. Was the market also highly fragmented? That's often like the Keith Raboy lens that he likes to look at. Both low NPS and. Because if it's low nbs, but there's just one monopolist, it might not be a good business opportunity. But what was that market map? How dense was it?

[Jubin Murzadogan]

Yeah, so I think the mother of all tailwinds for us right now, which is the challenge that every incumbent is going through, is that every 10 years there's a new CPQ vendor that that that comes alive. And the reason for that is because about every 10 years there's a new technology shift that happens. We went from software to SaaS that became a subscription based billing. Then we went from SAS to AI. AI is very likely going to be a usage or consumption based billing. Every time that happens and there's a new pricing pressure that comes on onto a business or a pricing model, you need somebody that can actually price. Like if you're a rep trying to do a deal, you need to be able to price that deal. If the data models with the incumbents were not built to be able to actually price Those deals, you're screwed. So every 10 years we have a new technology shift every 10 years that puts a bunch of pressure on pricing models. And every 10 years you need to be able to actually get quotes out the door leveraging those new pricing models. And so I think in our case, we are the beneficiaries of that.

[Host 2]

What is the sales process like? Because you mentioned the CIOs, the Chief Information officers, they feel like they have an incredible amount of leverage over the decision of what platform to use. But then the end user is different. So can you sort of walk through the actual user journey a little bit?

[Jubin Murzadogan]

Yeah, I would say the two primary Personas are CIOs who are the ones that are responsible for delivering software to an organization, and CROs who are the ones that are responsible for getting deals done inside of an organization. And the underlying software that they're using is just fundamentally broken. So those are like the two people that really matter. The unique thing about this problem is that there's many hands in the cookie jar, which is why it's like a unique problem for a startup to go solve because you have deals desk and rev ops and finance and sales and it all of these folks touch this business process. And so it's actually quite difficult both to build because you have to build for all of these Personas and sell because you have to get all of these people kind of on board with you. Even if you can get all of that done, then you have to convince them that running one of their most important production systems onto an early stage startup makes sense. And so I think that's like kind of the hill that we, I guess that we have the honor of climbing.

[Host 2]

Jordy.

[Host 1]

Yeah, just you're like, I'm pretty good at sales. Why don't I sell something that requires buy in from every part of the organization? Yeah, no, it's a great, it's a great challenge, but it seemingly will be incredibly sticky once you get everyone bought in.

[Host 2]

Where is AI good at this? Out of the box where foundation models are open source useful versus like you got to go build a harness or you got to go write some SaaS on top of it or get some of your own data and fine tune it. Like where are we on the front end tier of like this problem being solved end to end by AI?

[Jubin Murzadogan]

Yeah. So during the the series A fundraise, the, the hottest question was why won't anthropic eat you? Basically, and my general point of view is if they go after this, we're all screwed. Like we might as well put all of our money and all of our eggs inside of the anthropic basket because this feels like the most esoteric problem that you could possibly go after. I think the, the unique thing about this is, you know, at least at kp, we invested early in companies like Windsurf and Harvey and we saw what happened. When you can point these models at structured and unstructured text in nature, and anytime you can do that, the models are very, very good at reasoning with them. And CPQ is a very similar problem. You have price books, approvals, volume based discounting. All of these rules sit somewhere and the models are very good at reasoning with them. The challenge for us, I think, like where our kind of secret sauce is, is that you need to be able to have this agent architecture call it at the header. Then on the Y axis you have a bunch of policy engines that are enforcing a probabilistic system through a deterministic engine that sits on top of a data model that has to be flexible enough for the pricing models of today and in the future. So the combination of those three things is really our secret sauce. And then obviously, I think the problems that many of these kind of bleeding edge enterprise AI agent companies are running into is like, how do you get an agent to work predictably with the harness around it inside of a large enterprise doing something where if we go down, like we get sued, like you can't get quotes out the door and so you can't not get it right. And so I think, you know, in many ways we are tackling probably one of the more bleeding edge agent problems in the enterprise. And that's really where we live, is in the enterprise.

[Host 2]

Awesome. Tell us about the round. How much did you raise? I want to hit the gong.

[Jubin Murzadogan]

We, we raised 27 million in total.

[Host 1]

There you go.

[Jubin Murzadogan]

We announced the seed from KP 5.2 mil and then 22 million from. With founders fund leading and KP doubling down.

[Host 2]

Any familial conflicts going on over there? Just kidding. Trey Stevens led the round. Don't worry, he's brothers with a mean. Thank you so much for coming.

[Host 1]

Oh, I didn't, I didn't even put that together.

[Host 2]

Yeah, yeah, yeah.

[Host 1]

Amazing. Anyway, family business. Love it.

[Host 2]

Have a good one. We'll talk to you soon.

[Roman Chernin]

Thanks guys.

[Host 2]

Goodbye.

[Jubin Murzadogan]

Thanks guys.

[Host 1]

Talk soon.

[Host 2]

Up next, we have Roman Chernin from Nabius. He's the co founder and chief business officer. If you've been living under a data

[Host 1]

center and Nabius is only up 15% today, so we'll ask him.

[Host 2]

The AI cloud division posted 841% year over year revenue growth, earnings per share beat expectations significantly. Shares are up as you mentioned and the business has been on an absolute tear. And we're excited to be joined by Roman to break it down for us. How are you doing?

[Delian Aspro]

Great.

[Roman Chernin]

Thank you for inviting. I must say that I think three meetings ago.

[Host 2]

Yes.

[Roman Chernin]

Someone said that I need to take a nap so don't be surprised that I just fellow.

[Host 2]

Okay apart here, I can imagine you're working nonstop right now. How long have you been working on this? I feel like Nebius is in the news every day. But can you take us back a little bit on your journey with this company?

[Roman Chernin]

Yeah. So I'm with Nambius from the very beginning and Nabios officially in the current shape of the company exists from summer 2024, so less than two years. But we as a team had experience longer than that. You maybe know that core team came out from Yandex which was a large Russian Internet company. And yeah, we started with quite a unique mix of talent that helped us to start like really fast.

[Host 2]

So I mean people know Nebias Neo cloud, but that can mean a lot of things. How are you describing the shape of the business? How are you describing the.

[Host 1]

Before we get into that, what's the other spin out that's also on a tear from Yandex. Yandex, yeah.

[Roman Chernin]

You may be know Clickhouse.

[Host 2]

Clickhouse, that's right. Yeah.

[Host 1]

The talent bench at Yandex seems to have been just like genuinely insane.

[Roman Chernin]

You can find a lot of other startups or companies that created with Xiandex people. Yeah, not really mafia. Talking about how we describe our business, I don't really like the word new cloud because I think it's too broad definition of too many companies that do different things. Like someone is doing just really data center business. Someone is doing data center plus renting out hardware. Someone don't own hardware and do kind of marketplace play. We like to say about us that we built vertically integrated full stack AI, specialized cloud, AI specialized cloud. So that's what we think we do. And I think that the difference of Neo Cloud or bare metal compute and cloud is quite significant because it's completely different developer experience. Right. So if you need to come and rent a large cluster and wait like six months to be delivered and then sign it for three years and run it, probably it's not really real cloud experience. Right. In the cloud you expect more flexibility, faster time to value, more tools that you can use. As a developer, more layers of the value.

[Host 2]

How do you think about lagging edge model inference load over time? Because a new model comes out, there's an incredible amount of hype. Everyone's looking at benchmarks and testing it and then at a certain point open source catches up. Different models sort of commoditize. But if you're an enterprise and you found that GPT4 is great for scanning your receipts or something, you might leave that capability, that functionality running forever and you might never need to deploy a smarter model against that problem. And I'm wondering, as we see skyrocketing frontier revenues, lots of debate over token maxing versus incredible results on the profitability side, how are you trying to understand what's happening with those workloads that have been sort of baked into the economy already? Like the AI diffusion story is done because that capability has, has been adopted fully.

[Host 1]

Yeah.

[Roman Chernin]

So I think there are as we see, a few brands, let's say. So first, as you said, open source models are catching up different models quite, quite fast probably. I mean I think that like there is three, six months gap like that we see for the most of the tasks and we see that a lot of people, a lot of customers, when they come to the scale. So the journey is you start with the frontier model, you want to have the best capabilities to unlock the use case, to see that things working, to start growing. But then when you grow, you may meet the limitations. And first of all in economics, because at scale the economics matter. And then if you have other capabilities from open source models that can deliver in a particular use case that you already understand, like same or close to same performance and quality for times lower cost, you're actually switching and it's not easy to switch because you need to cook those models. Nobody uses obviously vanilla models. You need to like, you need to figure out your data and so on, so forth. But again, if you scaled, you have a lot of value there. And then what you need from infrastructure where we believe that our value is created, is to help you with that, to lower the barrier, to A, tune those models and then B run those models reliable and cost efficient. And that's like the layer, the next layers of the offering that they're building

[Host 2]

as you look out over the next couple years or even a decade into the future, which sounds like an infinite amount of time, but on the near and midterm, how are you thinking about semiconductor bottlenecks versus energy bottlenecks? So we've been going back and forth on this. Both are in short Supply. What's keeping you up at night?

[Roman Chernin]

I'm lucky person. I'm more on the product and demand side and it's a lot of fun. So I really, I. Yeah.

[Host 1]

Being that, being the demand guy during the explosion. Demand has got to be great.

[Roman Chernin]

The best worker.

[Host 1]

Yeah.

[Roman Chernin]

The best work in the world.

[Host 1]

Yeah.

[Roman Chernin]

So people hug you. But. But no, I think that now obviously the physical kind of the physical infrastructure, the even not the energy itself, but like getting from greenfield to having data center that works and full of GPUs is actually the big, the big challenge. And like the bottlenecks are like along the all the supply chain starting from whatever connecting to the grid or bringing local generation and all the way to like even human power and all the, all the complexity. So I really kind of take my head off from like in front of the people who are breeding those like gigawatt capacity online. And that's challenging work.

[Host 2]

Can you.

[Host 1]

How do you there. There's such overwhelming, overwhelming demand and just energy to bring compute online. How do you measure yourself and the team's performance? How do you know whether or not you're doing a truly great job? Right. Because people are so desperate that, that they'll work with like a third tier supplier just because it's their only option. You guys have shown tremendous capability. But then again it's like how do you know if you're actually doing an exceptional job?

[Roman Chernin]

It's a good question. I think the race is so that you never know are you good enough or not? So you look like, oh, you're growing like whatever, growing like seven times a year or so. Is it good enough like in any normal business? Yes, probably. It's excellent growth and you meet your customer and they say oh you know what, I grew seven times in the last four months. And like okay, probably, probably we need to.

[Host 1]

I gotta grind harder. Yeah, yeah.

[Roman Chernin]

But there is another component that I think is people speak less about which is how you finance all that. So. You asked about like the bottlenecks in the chips. You asked about like the bottleneck to bring physical infrastructure online. But then there is another part you need to finance all that. So I think that now if you would have unlimited capital we would be able to on technical execution, on operational execution, we could even move faster. So there are so many components in this business of AI infrastructure that you need to do efficient and financing, finance enough and finance smart to not eat all your margins. It's a part of the science here.

[Host 1]

What is a job at Nebbyus that is sort of the unsung hero of the business besides you. But we're singing for you now.

[Host 2]

He gets sung for all the customers

[Host 1]

are sitting, they're hugging you.

[Roman Chernin]

You said yeah, yeah, like.

[Host 1]

But is it like somebody that's like, yeah, boots on the ground.

[Roman Chernin]

Yeah. No. I think that our business is very much like at any cloud business to be honest. It's post sales business. So like when you sell, you give a promise, people hug you that you allocated capacity to them and like they had to start working with you and so on. But then how you make sure that the customers are happy and things are working. And again, everything is moving so fast. We, we've got new chips every three months, we got new physical data centers every month, we got new customers that of the scale that we never saw before every month. And the workloads are changing so fast. And I think that it's not like one role, I think that underappreciated but in general delivery in this and I started with that it's execution business. It's not that it's less magic. It's a lot of boring execution. Each layer starting from finance, physical infrastructure, software, customer kind of facing engineering support, every piece and then only magically everything comes together from supply chain to customer facing person that makes sure they don't waste their money with us. Then things start working. So yeah, I think delivery is the key. Delivery chain is the key in this business.

[Host 1]

Post sales.

[Host 2]

Last question. Can you talk about some of the recent acquisitions? Acquisitions and talent moves at Nebia. So I'd love to know the philosophy and strategy there.

[Roman Chernin]

Yeah, the philosophy is very simple. We need to build so many things that we need to move and we need to move so fast that we're always looking for people who can accelerate us. And it should be exceptional talent and it's. And or it should be something that has a great adoption. Our two recent acquisitions that we announced, just like with the two weeks, pose. Two teams that work on inference optimization. So you can think again. Big part of our business is how efficient we convert GPUs into tokens and value for the customers. And these are two teams. One of them again clarify AI one is very much focused on model optimization, like the engine of inference, how you run specific model and all the techniques around spec decoding, monetization and so and another is more like system optimization. All the routing, kiwi caching, orchestration across the big cluster of compute and so on. And we had a very strong internal team also working on inference. But we felt that we need to move Fast. Like we need to move faster. We need to bring more capabilities there because the market is so fast that every, every three months you can lose the pay. Like so, so big PI. So big PI. Piece of the pie.

[Host 2]

And it's going to show up in your, in your margins and your earnings obviously, which are important. Now if you, if you don't, if you have unoptimized chips, you're not going to get the yields that you want. The results and the tokenomics are going to flip.

[Roman Chernin]

Yeah, that's true. And also what's important that it unlocks the new types of the customers, the new types of the workloads that we can serve. Like not just against raw compute, but serve all this fast growing vertically companies and enterprise adoption.

[Host 2]

Yeah, that makes a ton of sense. Well, thank you so much for taking the time to join us on a busy day.

[Delian Aspro]

Yeah.

[Host 1]

Are you going to get some sleep or you'll sleep in a few years?

[Roman Chernin]

I will take a nap now.

[Host 2]

Fantastic.

[Host 1]

Just a nap. I know you're across the pond, right. So you should be just going to sleep for eight hours.

[Host 2]

But he said a nap.

[Host 1]

I love.

[Host 2]

Well, thank you for coming on the show, Roman. We'll talk to you soon. Have a good one. Goodbye. The last post I want to talk about today is about video games. Of course there's an individual who beat. Is this a Dark Souls boss. But they built a controller that uses their full body. It's not a VR game, it's not a VR simulation. But if you want to attack, you strike the dummy with a physical stick and that presses the button in the game to trigger your character to attack in the game. And then it appears that he's playing on a green screen or something because you can see the game actually playing behind him. And he also is wearing glasses of some sort. You don't like this? What's not to like? You need to get a workout while you're gaming. The future of gaming very well might very well produce some of the fittest people.

[Host 1]

Yeah, I don't like the aesthetics of this.

[Host 2]

You don't? You think you should. You're more in there.

[Host 1]

I shouldn't knock it till we try it.

[Host 2]

You're more. You're more a fan of.

[Jubin Murzadogan]

I've seen it.

[Host 1]

Let's get.

[Host 2]

You're more a fan of unironic larping. Like live action role playing where you get the. Like the SD Kid concert. You're a fan of that. Put on the actual armor. Do the real battle.

[Host 1]

Yes.

[Host 2]

Don't play the video game version of that. That's a poor simulacrum.

[Host 1]

Yeah.

[Host 2]

You want hit the park or find a multiplayer version of that? Just get another human to stand next to you and you can learn defense.

[Host 1]

Before we go.

[Host 2]

What?

[Host 1]

There's some. There's more of a scuffle on the timeline around the Meet GC campaign. I realize why Mark is taking shots at it, and it's because I think they use an actor to try to look like him. Let's watch it again.

[Host 2]

Oh, really Interesting. I did notice that the actor that plays the General Catalyst character is like over the top handsome in the sense that like the original Apple Mac and PC ad, Justin Long, he's sort of like an everyman. And the actor who's playing the General Catalyst guy looks like a male model to me. I don't know. You can be the judge of it. But yeah.

[Host 1]

All I'm saying is I think we got a Drake Kendrick situation on our hands. I think the new media team at Andreessen needs to respond. They need to try to dunk on General Catalyst gc. Okay, I want to see a war. I want to see. It should be a bloodbath anyways.

[Host 2]

Demanding acrimony in the tech industry. Sewing dissent.

[Host 1]

Yeah. No, no. They need to get revenge. What about they need to get revenge? Mark says this is a clear inversion, clever inversion of the original Mac versus PC commercials. This time clearly intended to make the sponsor appear smarmy and judgmental. Fascinating.

[Host 2]

You got rage baited. They rage baited you. You could have just muted this. You could have blocked them. You could have not amplified it. Now it's on everyone's timeline. How people war now because of this,

[Host 1]

we're going to get a VC content wrap up.

[Host 2]

No one would know what General Catalyst is until now. Now there's a fight and it's working. No, just kidding. It is entertaining, though.

[Host 1]

Very fun.

[Host 2]

But wait.

[Host 1]

Duke it out.

[Host 2]

Good campaign.

[Host 1]

It's time.

[Host 2]

I. I do. I do.

[Host 1]

Eric Torrenberg.

[Host 2]

Reggie James first. Oh, yeah. That's a matchup there. It is just, I mean, to, you know, put aside the war. It is just interesting seeing a produced 32nd, 62nd AD for a venture capital firm. It is sort of a first. I can't.

[Host 1]

I'm just happy to see a non vibreal video.

[Host 2]

Yeah, the vibrio meta that was sort of played out at this point. So we gotta go a different direction. And we're going to TV ads.

[Delian Aspro]

We're going to television.

[Host 2]

Full 30 seconds. Highly produced,

[Delian Aspro]

running on the super bowl

[Host 2]

on Saturday and live. I don't know. When would the VC advertise? Bloomberg cnbc? I don't know.

[Host 1]

We gotta get on with London.

[Host 2]

Yes.

[Host 1]

So we will see you tomorrow.

[Host 2]

Leave us five stars on Apple Podcasts and Spotify. Sign up for the [e-mail] and we will see you tomorrow at 11am Pacific Sharp.

[Host 1]

We love you.

[Host 2]

Goodbye.

[Roman Chernin]

Goodbye.