The Secret Economics Controlling OpenAI, Anthropic, and the AI Boom

Watch on YouTube ↗  |  February 18, 2026 at 00:30  |  22:50  |  TBPN
Speakers
John Coogan — Co-Host, TBPN
Jordi Hays — Co-Host, TBPN
Dario Amodei — CEO, Anthropic

Summary

John Coogan and Jordi Hays discuss the Cournot equilibrium shaping the AI arms race. They explain how a small number of AI labs compete on compute supply rather than price, analyze AI lab P&L dynamics between training and inference, and debate whether the industry settles into a durable hyperscaler-like oligopoly or shifts to Bertrand competition. The episode includes a Dario Amodei clip on AI lab economics and reviews OpenAI's recent Cerebras, orchestration, and enterprise moves.

  • Hosts explain Cournot equilibrium and apply it to the AI lab market.
  • AI labs are described as competing on supply and compute rather than price.
  • Training is a depreciating, cash-intensive game of chicken; inference margins are seen as healthy.
  • Dario Amodei clip covers AI lab gross margins and training scale-up losses.
  • OpenAI's Cerebras deal, orchestration hiring, and Frontier product are discussed as vertical integration moves.
  • Anthropic's Claude Code is cited as product leverage beyond an API business.
  • The end state is debated: durable oligopoly/hyperscaler-like vs commoditization/Bertrand competition.
  • A Nash equilibrium movie clip is shown as a thematic aside.
Ideas
John Coogan Co-Host, TBPN 1:01
AI labs are durable Cournot oligopoly.
The AI lab market is a Cournot oligopoly: a small number of labs compete on supply (data centers, GPUs, frontier and low-latency capacity) rather than price, keying off each other's compute investments. Because true frontier models have few substitutes and customers will pay up for the best model, prices stay high for frontier access. This supports a durable, hyperscaler-like oligopoly with healthy inference margins rather than perfect competition or pure commoditization.
AI-SECTOR HIGH
Jordi Hays Co-Host, TBPN 10:51
Claude Code gives Anthropic product leverage.
Historically Anthropic was criticized for being stuck on a hamster wheel of training the best model because its business was mostly API, where customers can swap to a smarter model. Claude Code gives Anthropic a product with more leverage, reducing the need to always train a marginally better model and making its business more durable.
ANTHROPIC MED
John Coogan Co-Host, TBPN 13:15
OpenAI moves strengthen vertical integration.
In a differentiated, vertically integrated AI market, OpenAI's recent moves make sense: Cerebras for speed, hiring Peter from OpenClaw for orchestration and model routing, and OpenAI Frontier for enterprise deployment. These moves deepen enterprise hooks and pricing power, positioning OpenAI well versus other labs.
OPENAI MED
John Coogan Co-Host, TBPN 13:52
Cerebras offers speed, validated by OpenAI.
Cerebras offers a differentiated product around speed. OpenAI's deal with Cerebras validates this speed edge, which should be valuable in latency-sensitive AI inference and in a vertically integrated AI market.
CBRS MED
Up Next
1:30:18
Stocks Gain as Tech Holds Up; Bonds Steady | The Close 2/17/2026
Bloomberg Markets · Feb 17
1:09:15
“Unrealized” Capital Gains Tax is Economic Suicide | WAYT?
The Compound News · Feb 17
5:48
Palo Alto Networks CEO Says Markets 'Have It Wrong' About Earnings
Bloomberg Markets · Feb 17
2:20:26
China is DOMINATING the US in EVERY statistic. Here's why. | Market Updates & News (LIVE)
Thread Guy · Feb 17
48:12
잘못된 건강 속설, 바로 잡아드립니다 | 윤방부 연세대학교 의과대학 명예교수 [더 헬스]
3PRO TV (삼프로TV) · Feb 17

This TBPN video, published February 18, 2026, features John Coogan, Jordi Hays discussing AI-SECTOR, ANTHROPIC, OPENAI, CBRS. 4 trade ideas extracted by AI with direction and confidence scoring.

Speakers: John Coogan, Jordi Hays  · Tickers: AI-SECTOR, ANTHROPIC, OPENAI, CBRS