Jun 21, 2025 · 2h 20m · tbpn
Weekly Recap: Elon's Starship Explodes, OpenAI's Defense Contract, Nvidia's Robots, We Test Cluely
John Coogan · 49m spoken Jordi Hays · 20m spoken George Hotz · 18m spoken Katie Haun · 14m spoken Justine Moore · 12m spoken Roy Lee · 5m spoken
0:00 / 0:00
▶ Watch on YouTube →
gold bands on the timeline = statements, start to end. Hover to read, click to jump. CC turns on captions
TBPN hosts John Coogan and Jordi Hays break down major headlines in aerospace, defense tech, and artificial intelligence, featuring live product evaluations and in-depth interviews with industry leaders Katie Haun, Justine Moore, George Hotz, and Cluely founder Roy Lee.
How this conversation actually went
Every chapter scored 0–10 on four independent dynamics. Hover any point for the reasoning behind the score. The hosts hold 56.3% of the talking time here. How this is scored →
The hosts as informed peer 2.5 Guest teaching 2.2 Guest disagreement 1.7 The hosts pushing back 1.8
speaking balance: gold is the hosts, purple is the guest (3 minute bins)
Sharpest disagreement ▶ 2:04:27 Hotz mocks factory humanoid robotics
George Hotz forcefully attacks the humanoid robotics narrative, claiming anyone advocating them has never worked in a factory and daring anyone to show a single arm that can autonomously insert a loose screw.
Hardest push from the hosts ▶ 20:59 Coogan challenges Roy Lee on mobile ecosystem lockoutJohn Coogan refuses Roy Lee's easy dismissal of mobile interfaces, pointing out that Apple's strict privacy sandboxing will block Cluely from the majority of consumer mobile usage.
Biggest teaching moment ▶ 1:00:50 Haun breaks down market structure vs stablecoin legislationKatie Haun educates the hosts on why settling for only a stablecoin bill is an industry trap, framing the market structure bill and the Supreme Court's Chevron doctrine reversal as the real battlegrounds.
The host holds their own ▶ 21:38 Coogan drills Roy Lee on Meta Ray-Bans vs Apple ecosystemJohn Coogan demonstrates detailed domain knowledge by contrasting Meta's open hardware API ecosystem with Apple's closed walled garden, forcing Roy Lee to pivot his strategic argument to desktop enterprise value.
the scores for every segment, with the reasoning behind each
| Chapter | Topic | The hosts as informed peer | Guest teaching | Guest disagreement | The hosts pushing back | Why |
|---|---|---|---|---|---|---|
| SpaceX Starship Explosion and the Challenges of Orbital Scaling | 0 | 0 | 0 | 0 | Monologue segment where the co-hosts analyze the SpaceX Starship explosion and explain orbital window mechanics and industrial scaling economics. As there is no guest, all interaction scores are zeroed per scoring rules. | |
| Meta's $5 Billion Investment in Eyewear Giant Luxottica | 0 | 0 | 0 | 0 | Hosts discuss Meta's investment in Luxottica, OpenAI's defense contract, and WhatsApp advertising monetization. There is no external guest present. | |
| Live Interactive Demonstration of Cluely AI Assistant | 3 | 1 | 0 | 2 | Co-hosts test intern Tyler in real-time with trivia queries to evaluate the Cluely AI assistant. Tyler cooperates fully while reading streaming results from the tool. | |
| Productivity Tool Sponsorship and Cluely Capability Testing | 2 | 1 | 1 | 2 | Continuation of Cluely testing alongside a Linear sponsor integration. The hosts lightly tease Tyler about cognitive reliance on the tool. | |
| Interview with Roy Lee on Cluely's $15M Series A and Strategy | 6 | 3 | 7 | 6 | Roy Lee aggressively hypes Cluely's $15M raise, dismissing conventional VC processes and phone form factors. Hosts challenge him on iOS platform restrictions and ask whether he is building a durable product or personal fame. | |
| Mark Walter's Historic $10 Billion Los Angeles Lakers Acquisition | 0 | 0 | 0 | 0 | Co-host recap of Mark Walter's $10B purchase of the Los Angeles Lakers and sports franchise ROI versus index investing. Monologue segment with no guest. | |
| The Data Labeling Ecosystem: Surge AI Outpaces Scale AI | 0 | 0 | 0 | 0 | Host analysis of Edwin Chen's bootstrapped Surge AI outpacing Scale AI in revenue following Meta's quasi-acquisition. Monologue segment with no guest. | |
| European Defense Tech Acceleration and Helsing's $600M Financing | 0 | 0 | 0 | 0 | Discussion of Helsing's $600M funding round led by Daniel Ek and European defense tech scaling. Monologue segment with no guest. | |
| Katie Haun on Stablecoin Regulation and Crypto Market Structure | 6 | 7 | 2 | 3 | Katie Haun systematically educates the hosts on why the market structure bill matters far more than the GENIUS stablecoin act alone. She explains how the Chevron doctrine overturn shifts power from agencies to courts. | |
| a16z's Justine Moore Maps the Generative Video Landscape | 6 | 6 | 1 | 2 | Justine Moore maps the generative AI video landscape, breaking down vertical tools, dense captioning quality, and Chinese model competitiveness. The dynamic is collaborative and informational. | |
| Nvidia and Foxconn Deploy Robotics for Server Manufacturing | 0 | 0 | 0 | 0 | Hosts critique the marketing hype around Nvidia and Foxconn deploying humanoid robots for server manufacturing. Monologue segment with no guest. | |
| George Hotz on AI Scaling Limits, Compute Moats, and Robotics | 7 | 8 | 9 | 6 | George Hotz brings extreme contrarian combativeness, dismissing terms like AGI and GDP while dismantling humanoid robotics hype and tele-operated self-driving metrics. Hosts repeatedly probe and test his frameworks on compute moats, data efficiency, and diffusion models. |
Statements from this episode (34)
Coogan: Aerospace challenge is scaling industrial economics, not one-off technical feats
“The challenge now is not, can we get to the moon? It's the same thing with like, the challenge is not, can we build a flying car or can we build, we have helicopters. Can we build one humanoid robot or one self-driving car in San Francisco? It's like, can we a…”
Coogan: Meta is reportedly investing $5 billion for 4% of Luxottica
“Luxottica owns both Oakley and Ray-Ban and Meta is reportedly investing five billion dollars for a four percent stake.”
Coogan: OpenAI secured a $200 million contract with the US DoD
“They scored a two hundred million dollar US defense contract. So they're working with the DOD.”
Lee: Mobile phones may not be dominant consumer devices in five years
“Like I think phone, like in five years, I would question whether phone is the dominant consumer use case.”
Lee: Only 2-3 competitors are seriously building desktop AI assistants
“There's maybe two, three competitors in this, maybe, maybe two, three competitors who are actually trying to take a meaningful stab at a desktop assistant, and there's way more money and way more value in this than people think.”
Coogan: Los Angeles Lakers sold for $10B in richest sports deal ever
“The Los Angeles Lakers has been sold for ten billion in richest deal in sports history.”
Coogan: 40-year return on owning Lakers slightly underperformed the S&P 500
“And I looked at the return on investment of owning the Lakers for that 40 years. Slightly under S and P 500.”
Hays: Bootstrapped startups struggle for hype without venture investors pushing them
“It's so incredibly hard to have any hype around your business if you're bootstrapped because you're not having your investors aren't hitting the timeline for you on a daily basis. And also you have, if you're not trying to raise capital, you have less need to …”
Coogan: Surge AI generated over $1B in 2024, surpassing Scale AI
“The company generated more than one billion dollars in revenue last year. Surge has told employees, a previously reported figure that exceeds The eight hundred and seventy million dollars scale generated in revenue during the same time period.”
Coogan: Surge AI has been profitable since inception, unlike Scale AI
“And unlike scale, surge was profitable and has been from the beginning, Chen said.”
Coogan: Surge AI bills two to five times more than Scale AI
“Chen has distinguished Surge by making it the high-end shop, charging premium rates often two to five times what scale might bill.”
Coogan: Anduril partners with Rheinmetall to manufacture Barracuda and Fury in Europe
“Over in defense tech world Andrew has partnered with Ryan Mattall, the German the German defense tech company, the prime really to manufacture Barracuda, Barracuda and fury over there.”
Coogan: Daniel Ek leads $600M round into Helsing at $12B valuation
“So yeah, Spotify founder Daniel Eck is leading a six hundred million dollar funding round into the German defense tech startup Helsing. This is the, this is kind of more of an annual equivalent over in Europe valuing that four year old company at twelve billio…”
Coogan: Helsing expands from AI software into drones, aircraft, and submarines
“Helsing is, is expanding from its origins in artificial intelligence to produce its own drones, aircraft, and submarines as part of the bigger push for locally made and owned defense products for European countries and really companies all over the world.”
Haun: The future will not involve every major company issuing a stablecoin
“Is there going to be a world where every company has their own stable coins? Yeah. We don't think so. We could be wrong, but that's not the view of how we see this evolving.”
Haun: Crypto market structure bill would have prevented past industry blowups
“The irony is if you would have had the market structure bill, you wouldn't have had a lot of the blowups that you had. In the past several years in this industry.”
Haun: AI creates digital abundance while blockchains enforce digital scarcity
“AI creates digital abundance, and blockchains are good at enforcing digital scarcity. And I think you're gonna see more and more synergies emerge and use cases over time”
Haun: Gensler targeted compliant crypto firms while ignoring complete disasters
“Gary Gensler basically picked what should have been the best companies in crypto, the poster children for compliance, and brought enforcement actions against them, and then the complete spectacular disasters didn't bring anything, so until after the fact.”
Moore: ByteDance's SeedDance video model is already outperforming Google's Veo 3
“The new Minimax model, the new ByteDance model, SeedDance, which is in the arena, already outperforming VO-three.”
Moore: OpenAI scrapes YouTube to train its artificial intelligence models
“I mean, we've seen companies like open AI will scrape YouTube as well to train their models.”
Moore: China has a video AI advantage from cheaper PhD labelers
“And China has really benefited there because there are so many more PhDs than there are here, and it's much cheaper for these companies to hire them to do these dense labels for the video data.”
Moore: ElevenLabs offers an upload tool to detect its AI-generated audio
“They actually have a site where you can upload any audio and it will tell you if it was generated with 11 labs or not.”
Hays: Manufacturing is likely not a great use case for humanoid robots
“We've had, we've asked a bunch of robotics experts about humanoids. And so far, I haven't got a, I don't have the confidence that, that this is Actually a super great use case for them.”
Coogan: Unitree robots can execute specific tasks repeatedly, though not yet generalizable
“Unitry is, is already like, feels like it's scaling up to the point where like unitry robots should be usable in certain locations. Like, they're not generalizable yet, but they're certainly, if you train them on a specific task, they could do that task over a…”
Hotz: AI is a far better military technology than nuclear weapons
“So, like, as a military technology, nukes are not that good. AI is way better.”
Hotz: Either Nvidia is overvalued or AMD is significantly undervalued
“Either Nvidia's really overvalued or AMD's really undervalued. It has to be one or the other.”
Hotz: AI application wrappers like Cursor and Windsurf will fail
“And then on top you have, like, completely worthless things like cursor and windsurf you know, these character AI, all these people who think, oh, we're the app, we're gonna get the ARR. No, that worked in the web, it won't work for AI”
Hotz: AI scaling laws deliver linear returns for exponential capital
“AI scaling laws. You can put in exponentially more money to get linear returns.”
Hotz: Waymo robotaxis require approximately 1.2 remote human operators per car
“There's probably about, I would say there's 1.2 operators per car. But it's not, they don't have a steering wheel and pedals. It is an autonomous system that they're probably doing some higher level inputs on.”
Hotz: Robotaxis will be a scooter-style capital race to the bottom
“Self-driving cars look like scooters. The only thing that it's going to take to roll out big fleets of self-driving cars is capital, right? It's just strictly a capital market... So self-driving cars are going to be this awesome race to the bottom, right? It's…”
Hotz: Robotic arm hardware is good enough, robotics is strictly software
“The arms are already good enough. It's the off the shelf arms are fine. It's all software. Again, it's always all software. Autonomous vehicles are all software. Robotics is all software”
Hotz: Current machine learning models are 1,000x less data efficient than humans
“Like current machine learning algorithms, like a thousand X less data efficient than humans. So yeah, you need a thousand X more data, right? If a human can learn something in one example, or 10 examples, the computer is going to need a thousand or 10,000.”
Hotz: Automated flight booking will arrive via APIs, not agentic web browsing
“We're gonna pretty soon have computer use models that are actually capable of going to Delta.com and booking a flight. But then what's actually gonna happen is Delta's gonna partner with whatever company does that, and they're gonna put it behind the stupid th…”
Hotz: Shift to diffusion AI models will erode centralized cloud moats
“With AI, I think there's going to be a much less of a moat. Especially when you look at the move from autoregression to diffusion. So autoregression can run in large batch sizes. When you run ChatGPT, you're running with a whole bunch of other people on that s…”