Demis Hassabis proposes a FINRA-style federal AI standards body while Apple clears China with Alibaba and Baidu models — and Moonshot's Kimi K3 narrows the frontier gap. Today's briefing maps how governance is becoming a competitive weapon across every major market.
Audio is available on Spreaker — see link below.
Demis Hassabis wants the US government to approve AI models before they ship. That's the sharpest version of what the Google DeepMind CEO proposed this week, calling for a federally-overseen public-private standards body modeled on FINRA, the financial industry regulator.
The market's counter-argument is speed. Any mandatory review window creates a gap that competitors, including state-backed labs, won't observe.
Apple cleared China's Cyberspace Administration after twenty-two months. Apple Intelligence is now approved in the world's largest smartphone market, but the terms matter.
Moonshot, the Chinese AI startup, launched Kimi K3 this week. It's an open-weight model with two point eight trillion parameters and a one-million-token context window, and it's performing competitively with Anthropic's Fable on benchmarks.
Fireworks AI closed a one point five billion dollar Series D at a seventeen point five billion dollar valuation. The company processes forty trillion tokens daily and is running over one billion dollars in annualized revenue.
In Japan, Nvidia's Jensen Huang announced the Noetra partnership with Sony and Honda, anchored by a twenty-seven thousand five hundred GPU national AI super factory. The focus is physical AI and robotics.
The thread across all of this is governance as leverage. Hassabis wants to define the standard.
Chapter summary auto-generated from the verified script. Listen to the full episode for the complete content.