Paradox lost
Computing history has run the Jevons Paradox experiment at least 28 times. Demand won every time. The capital lost anyway.
James Thomason
Notes on investing in technology disruption. Engineer → Founder → CTO → VC & long/short equity PM. 25 years of Silicon Valley pattern recognition.
Latest posts
Open models, closed fist
After enclosing the social web, Zuckerberg warns that concentrated power is dangerous. Now he tells us.
The value chain cascade
Why the most important phase of a technology begins the moment it stops improving.
It's shovels all the way down
Memory, guardrails, compression, identity. The whole ecosystem of AI shovel-makers is about to go extinct, killed by the platforms who sell the ground they stand on.
Google, a $100B SpaceX shareholder, suddenly needs $11B of SpaceX compute before IPO
Google calls it a short-term bridge for Gemini. The bridge happens to prop up a $1.75 trillion valuation.
Boxabl folding home SPAC is an unfolding disaster
$1.5M in revenue for 23 boxes delivered, $200M burned, a meme coin, related-party transactions, and 50,000 investors who can't sell into a nosebleed $3.5B valuation
Nasdaq drains 401ks into SpaceX, Anthropic, OpenAI IPOs
In which the index gods rewrite scripture so the faithful may be sacrificed on schedule.
SpaceX's IPO Bagship carries full payload of Elon's mistakes
In which the great American space dream, it turns out, is just another vehicle where the greater fool flies coach.
Where bad ideas go to fly
SpaceX discovers the one thing AI was missing is rockets somehow
Narratives pick the KPI, and the KPI picks the losers
How the metrics that attract capital in emerging tech sectors systematically favor the companies least likely to dominate long-term
The pricing game versus the valuation game
Alpha for general purpose technologies skews later than you think
Too many winners, not enough market
A simple framework for identifying sectors priced beyond arithmetic reality reveals which AI layers face inevitable repricing
The option premium view of bubbles
Why we need our speculative manias.
Moore or less the bottleneck
What happens if software outpaces hardware in AI?
The AI utility trap
Rapid AI commoditization could be outpacing traditional value capture mechanisms