Now Playing
Secret Rulebook: White House AI Review, Qwen3 & Big Tech's Hidden Losses
Audio is available on Spreaker — see link below.
The White House is finalizing a frontier AI review framework — with classified standards the companies being judged can't see. Today's briefing covers the regulatory blind spots, Alibaba's open-source pressure play, Anthropic's Pentagon ban, and why Big Tech's AI earnings aren't what they appear.
Audio is available on Spreaker — see link below.
The White House is meeting today with OpenAI, Anthropic, Google, and Meta to finalize something that's never existed before: a formal government review framework for frontier AI models. Thirty days of pre-release access before a model ships.
There's also no designated White House leader for this review system. Three officials, Sean Cairncross, Scott Bessent, and Howard Lutnick, are reportedly competing for ownership.
Alibaba released Qwen3.8-Max this week. Two-point-four trillion parameters, one million token context window, and critically, it's open-source.
The Anthropic situation deserves its own frame. The Pentagon has designated Anthropic a supply chain risk, blocking military use over disagreements on safety guardrails.
Over twelve hundred AI executives, including Anthropic's Dario Amodei, signed an open letter this week calling on government to develop tools to pace AI development. The catalyst was a series of agent hacking incidents disclosed by OpenAI and Anthropic.
One more signal worth watching. Amazon's earnings were reported at two hundred and forty percent growth.
The two things to track from here are straightforward. First, whether Tuesday's White House meeting produces a framework with actual scope clarity, an assigned leader, and a definition of open-weight inclusion.
Chapter summary auto-generated from the verified script. Listen to the full episode for the complete content.