Feb 18, 2026 · 30m · tbpn
The Cournot Equation, Micron’s $200B Bet, Hollywood vs. Seedance 2.0 | Diet TBPN
John Coogan · 13m spoken Jordi Hays · 9m spoken Dario Amodei · 2m spoken John Nash · 32s spoken Martin Hansen · 21s spoken Dwarkesh Patel · 8s spoken Princeton Student · 6s spoken Podcast Contributor · 4s 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
The hosts analyze the Cournot economics and game theory driving frontier AI labs, alongside global semiconductor megaprojects and Hollywood's escalating battles against generative video models.
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 49.4% of the talking time here. How this is scored →
The hosts as informed peer 4.3 Guest teaching 0.7 Guest disagreement 1.1 The hosts pushing back 1.3
speaking balance: gold is the hosts, purple is the guest (3 minute bins)
Sharpest disagreement ▶ 27:57 Guest insists bad actors can outpace IP enforcement
Hays asserts that rogue ad accounts can successfully monetize deepfakes of prominent creators like Huberman before anyone notices, challenging conventional assumptions about brand protection.
Hardest push from the hosts ▶ 28:09 Coogan rejects Huberman ad clone hypotheticalCoogan explicitly counters Hays with a direct refusal ('I would disagree'), arguing that Huberman's team is too sharp to allow unauthorized clone ads to run undetected.
Biggest teaching moment ▶ 8:34 Amodei explains gross profits vs exponential model research costsIn the featured interview clip, Dario Amodei clearly articulates the distinction between 75% gross margins on inference and overall corporate unprofitability caused by $10B exponential training scale-ups.
The host holds their own ▶ 2:42 Host demonstrates market dynamics of AI inference and hardwareCoogan shows command of industrial economics by breaking down positive contribution margins in GPU inference versus open source commodity hardware benchmarks.
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 |
|---|---|---|---|---|---|---|
| Lunar New Year and Year of the Horse Banter | 7 | 1 | 1 | 1 | Coogan delivers an extended, highly detailed breakdown of Cournot equilibrium, Bertrand competition, and positive contribution margins in frontier AI inference. Hays chimes in with supportive color commentary about Anthropic's business model and Claude Code. The dynamic is collaborative and co-host banter. | |
| Dario Amodei on AI Lab Margins and Training Scale-Up | 0 | 0 | 1 | 0 | This segment consists of a recorded clip playback featuring Dario Amodei explaining gross margins versus exponential training scale-up spending to Dwarkesh Patel. Because this is a third-party audio clip, host activity is scored at zero. | |
| Game Theory and 'A Beautiful Mind' Clip | 4 | 0 | 1 | 1 | Coogan and Hays watch and comment on a scene from A Beautiful Mind. Coogan contextualizes the stylized depiction of game theory and Adam Smith's economic theories. | |
| Automating Knowledge Work and Video Editing Bottlenecks | 6 | 1 | 1 | 2 | Hays raises the Dwarkesh Patel point about software engineering having total context in codebases unlike other jobs. Coogan counters and elaborates using his own video editing experience, explaining why creative editorial decisions and cutting room floor choices cannot be captured like GitHub PR logs. | |
| Pop Culture Memes and Record-Breaking Girl Scout Cookies | 4 | 1 | 1 | 1 | The hosts cycle through news items including Girl Scout cookie records, Adani's infrastructure push, Micron's memory spending, and PlayStation 6 delays. Coogan discusses future shifts toward cloud-based gaming and generative up-res passes. | |
| Warner Bros. Resumes Acquisition Talks with Paramount | 4 | 1 | 1 | 1 | The pair discuss Warner Bros. M&A talks and the internet virality surrounding OpenClaw and Anthropic's trademark handling. The interaction is friendly and banter-driven without conflict. | |
| SAG-AFTRA and Hollywood Battle ByteDance's Seedance 2.0 | 5 | 1 | 2 | 3 | Hays covers SAG-AFTRA's response to ByteDance's Seedance 2.0 and speculates on ad account unauthorized cloning. Coogan directly pushes back on Hays's hypothetical about cloning Andrew Huberman, defending the Huberman team's vigilance. |
Statements from this episode (14)
Hays: End-user applications give AI labs more leverage than pure API businesses
“Having a product, not just an API business gives you leverage because at some point the models are smart enough where you don't need to train them. You don't need to train a model that is four percent better because people are still coming to your application …”
Coogan: Frontier AI will remain a long-term oligopoly of three labs
“And also it sort of explains why a lot of the VC firms are getting in multiple companies because they don't think it's going to be winner take all anymore. They think it's going to be much more oligopolistic for the longterm, and there will be competition betw…”
Amodei: Frontier AI oligopoly will not equilibrate to zero margins
“If there's like three firms, if there's three firms in the economy, all are kind of independently behaving, behaving rationally, it doesn't equilibrate to zero.”
Amodei: Frontier AI gross margins are currently very positive
“Yeah, so the, again, the gross margins right now are very positive.”
Amodei: Deployed AI models are profitable, but labs lose money scaling training
“Each model makes money, but the company loses money.”
Hays: Current AI tools cannot adequately edit videos or pick moments
“We've tried every tool there is. They can't do it yet.”
Coogan: AI video editing lacks GitHub-style public training logs
“One interesting thing is that the, there isn't, there aren't a lot of open source, like, premier profiles. Like, I, I've edited a ton of videos for YouTube. There's a whole bunch of cuts in there. What I cut out, what I didn't. You could have that record, but …”
Hays: Adani Group commits $100B to Indian data centers by 2035
“India's Andani Group and energy and logistics giant said it would invest a hundred billion dollars to develop large scale data centers by 2035, the largest such commitment in India so far.”
Coogan: Generative AI up-resing will drive future video game graphics progress
“In theory, you should be able to run, like, a Gen AI up-resing pass to make it more photoreal, and I feel like that's gonna be where more of the juice is squeezed out of the graphics than just continuing on the traditional path of, like, more pixels, more ray …”
Hays: Warner Bros. resumes talks with Paramount after two-month standoff
“He says this morning, Warner Brothers is going to resume talks with Paramount after two months of rejecting them, playing mind games.”
Hays: Dave Morin to lead OpenClaw foundation as creator joins OpenAI
“Yeah, Dave Dave Morin I remember reading, is gonna step into I believe run the foundation that will kind of steward the open source project, and then Peter's obviously joining OpenAI.”
Hays: Reports indicate Claude was used in Maduro raid planning
“There was reporting last week that Claude was leveraged in some way during the Maduro planning, the planning of the Maduro raid.”