June 15, 2026
AI Economics Is Broken, Part II
Falling AI token prices help customers and labs sell adoption, but they can also mark down the earning power of the GPU fleets sitting on cloud-provider balance sheets.
Field Notes
Current InflectAI field notes on beliefs, markets, narrative drift, and systems change.
June 15, 2026
Falling AI token prices help customers and labs sell adoption, but they can also mark down the earning power of the GPU fleets sitting on cloud-provider balance sheets.
May 30, 2026
The SpaceX/xAI S-1 makes it possible to compare public AI token pricing with the fuller cost stack behind frontier inference.
May 25, 2026
The Federal Reserve's January 2026 bank-lending survey shows how AI adoption pressure can become a capital-market signal before every CEO fully believes the technology story.
May 23, 2026
CoreWeave shows how equity holders, short sellers, credit investors, and customers can look at the same AI infrastructure company and see different facts through different professional lenses.
May 18, 2026
The Railway Bubble of 1847 shows how a real technological transformation can become financially unstable when novel instruments make demand look like proof.
May 14, 2026
Professional identity creates belief pools in financial markets: populations trained by role, incentives, vocabulary, liability, and measurement to process the same facts through different instruments of perception.
May 13, 2026
Telling an AI system not to make mistakes is categorically similar to telling a new hire not to make mistakes: it names the desired absence of failure without transferring intent, constraints, context, or verification.
May 11, 2026
Market participants do not update beliefs evenly because professional identity shapes what information becomes salient, credible, and actionable.
May 10, 2026
AI may be different because cognition has become cheap enough to distribute, but real technological transformation has never protected markets from overinvestment.
May 8, 2026
Beliefs have shape. Meaning has structure. Structural changes in meaning can sometimes be seen before they become visible in ordinary metrics.