The Information Machine

Quoting Dean W. Ball

Simon Willison · Simon Willison · 2026-06-26

Simon Willison highlights Dean W. Ball's argument that US AI export restrictions are economically self-defeating because frontier AI labs depend on a global total addressable market to recoup massive training costs within the narrow post-release window before models become sub-frontier.

Open original ↗

Appears in

Extraction

Topics: ai-policyai-regulationfrontier-modelsai-economicsexport-controls

Claims

  • Frontier AI models must recoup a significant fraction of their enormous training costs within the narrow window of months after release, before competition erodes margins.
  • US export restrictions shrink the total addressable market for AI services below what is needed to justify $100 billion-scale data center investments.
  • The ongoing AI infrastructure buildout assumes a functionally global market for US AI services, not a restricted subset of approved customers.
  • Every week of regulatory delay directly destroys commercial value for AI labs by eating into the finite post-release window for margin capture.
  • The US government's AI export policy is described as a bad state of affairs by Dean W. Ball.

Key quotes

Every week of delay is eating into the narrow window that labs have to make their accounting work.
No one is building $100 billion dollar data centers to serve frontier models to whatever 100 companies the US government will allow access.
The ongoing AI infrastructure buildout—the one that is, according to former US AI Czar David Sacks, essential to the US economy, assumes a functionally global total addressable market for US AI services.