Sep 14, 2026 · 13 min
AI’s safety race collides with Washington’s growth agenda
Asking the U.S. Government to Regulate AI
The episode frames AI regulation as a choice between public safeguards and an administration betting on rapid expansion to ease economic pressures.
- 1Dario Amodei warns that AI capabilities are advancing faster than engineers can understand their consequences.
- 2Public oversight and enforceable safety standards could counter risks from unsafe systems and automated programs overwhelming the internet.
- 3The Trump administration supports rapid AI expansion and data centers while hoping AI-driven growth can offset slowing productivity and rising deficits.
Don't miss
Dario Amodei’s warning that AI progress is outpacing engineers’ understanding becomes the episode’s case for enforceable public safeguards.
The brief
Michael Moss substitutes for Heather Cox Richardson in a September 13, 2026 letter examining Anthropic CEO Dario Amodei’s warning about AI’s accelerating risks.
Amodei argues that model improvements are outpacing engineers’ understanding, raising the prospect of unsafe systems and automated programs overwhelming the internet.
The episode turns that warning into a case for public oversight and enforceable safety standards rather than leaving AI’s trajectory entirely to private companies.
It then connects the Trump administration’s support for rapid AI expansion and data centers to hopes that AI growth can counter slowing productivity and rising deficits.
The episode’s key tension is whether economic urgency should accelerate AI deployment when the people building these systems say their risks are becoming harder to understand.
Featuring
Michael Moss
Anthropic
Donald John Trump
OpenAI
Google DeepMindListen to the full episode and explore every guest, topic, and moment on PodLume.