Opinion analysis with tracked predictions and deep dives
The reference class says winning a Warner Bros. bidding war is losing. Netflix let Paramount win. The deal only makes sense if you're not trying to make money.
The model that said "leverage" now watches a war without a defined end state. The only mission with a clear endpoint is nuclear material seizure. Everything else is open-ended.
The national debt is $36 trillion. The deficit hawks have been predicting the explosion for 30 years. The bomb is the wrong metaphor.
The U.S. has assembled more air power in the Middle East than at any point since 2003. It's leverage, not a war plan.
Netflix, Paramount, and the DOJ are fighting over Warner Bros. The prediction markets say it's a coin flip.
Freddie deBoer will win his AI wager with Scott Alexander. But "normal technology" is the most transformative force in human history — and he's testing the wrong timeframe.
Two carrier strike groups in the Middle East reveal a negotiation pattern, not a war plan.
Starmer's crisis isn't about Mandelson. It's about a government that won 411 seats with no idea what to do with them.
How Takaichi's landslide locked America into the China fight — and why Beijing's pressure campaign backfired
Legacy SaaS is becoming the new mainframe: still running, still collecting checks, but no longer where anyone builds anything new.
SaaS systems of record are becoming primitives. The real question is who captures value in the agentic layer above them.
Thesis momentum across demographics, financial repression, inflation, and reshoring
Stalin and Mao showed us the pattern. We just can't see inside to know where it ends.
Shutdowns don't change policy. Democrats know this. They'll take the political win and move on.
Mark Carney's Davos speech gave middle powers the rhetoric they wanted. It won't give them the autonomy they need.
Google's comeback may be more fragile than the stock reflects
The cheating debate is a sideshow. Students need to learn to build and work with agents.
A time traveler from 2027 explains why the skeptics are measuring the wrong thing
My AI rejected this thesis. Twice. Here's why it's wrong.