Moats Built for Humans
Software moats—workflow lock-in, interface training, data silos—were built to stop humans. AI agents bypass all of it. The $300B wipeout was the market catching up to a structural shift.
Represents the posts that are dedicated to understanding the implications of businesses and their strategy on the society
Software moats—workflow lock-in, interface training, data silos—were built to stop humans. AI agents bypass all of it. The $300B wipeout was the market catching up to a structural shift.
Everyone's arguing about who's winning AI. Meanwhile, OpenAI made $20 billion and lost $14 billion. Chinese startups burn cash even faster. The uncomfortable truth? Neither side knows how to make money from this yet. They're just losing it differently.
Volkswagen spent 5 years and billions of euros building software that never worked, then paid $5.8B to license Rivian's. The federated structure that enabled twelve brands systematically prevented building the unified system software-defined vehicles require.
SSI, Thinking Machines, World Labs, Humans&—each represents a different bet on AI's future. But ambition isn't the variable that matters. Scarcity is. The Great Filtration will separate those who control something defensible from those who don't.
BMW achieved 5.2% Q3 margins versus Mercedes' 4.8%, a 40-basis-point gap that either proves strategic transformation remains possible with better execution or merely demonstrates BMW hasn't yet hit the constraints that killed its rival's strategy. Four paths forward, one answer by 2028.