The Moat or the Commons

5 hours ago

U.S. capital financed AI with the assumption that frontier models would become a monopoly business, justifying massive investments with high, protected margins.
Open-weight models, largely released by Chinese labs and supported by Western open-source infrastructure, are commoditizing AI capability, reducing the performance and cost gap with closed models to six to twelve months.
The conflict between the expected monopoly and the reality of commoditization is shaping U.S. AI policy, as capital seeks to create artificial scarcity through regulatory, vertical, and distributional means.
Predictions for the U.S. direction include: regulatory enclosure framed as security against Chinese models, labs vertically integrating to become service operators, and a split market where the U.S. protects domestic margins while losing global share.
This protectionist approach mirrors the decline of the U.S. auto industry, where walls led to complacency, higher costs for consumers, and long-term uncompetitiveness.
The costs of protectionism fall on U.S. consumers, small developers, labs, and global influence, while benefits accrue to labs, cloud providers, capital, and politicians.
Recommendations include building on open commons, designing for jurisdictional flexibility, and accounting for policy changes in system architecture to navigate the evolving landscape.

Hasty Briefsbeta