Skip to main content
AmigoFish

What METR Found at Anthropic + Nvidia's Crisis of Trust

From The AI Policy Podcast · · 57 min

This plays from the publisher's own server. We do not host a copy.

Open at the source

Not rated yet

Description

This week, co-hosts Aalok, Nicole, and Chris break down the revelation that Nvidia CEO Jensen Huang had an affair with an alleged Chinese spy tied to Megaspeed. They also unpack METR's recent report on accelerated AI R&D at Anthropic, before returning to last week's midterm elections and the future of U.S. AI policy.

NOTE: This episode was recorded Wednesday November 11, before reports of Jensen Huang’s second affair became public, or the Wall Street Journal report that Nvidia’s board is considering ousting him.

Timestamps:

01:02 Nvidia CEO faces spy scandal

05:59 Legal remote access vs. illegal chip diversion

08:56 What the U.S. government knew, and when they knew it

16:39 Why the alleged spy connection matters

18:05 Policy implications

25:15 METR report on the pace of AI progress at Anthropic

34:12 Why RSI is a safety concern

37:15 Reactions from authors of AI 2027

43:23 U.S. midterm election results

50:53 What to expect from the new Congress on AI policy

55:17 Other news: Anthropic IPO + new Fable, Mythos 5.2 models Additional Reading: “Nvidia Chief Seen as Target of Chinese Spy Plot” (The New York Times): https://www.nytimes.com/2026/10/29/us/politics/nvidia-china-jensen-huang-trump.html

Jensen Huang’s statement to Nvidia employees: https://nvidianews.nvidia.com/news/a-message-to-our-company

AI chip imports by country (Epoch AI): https://epoch.ai/data/ai-chip-imports?view=graph&tab=h100_equivalents

METR’s assessment of AI R&D acceleration at Anthropic: https://metr.org/blog/2026-11-05-anthropic-ai-rd-acceleration-assessment/

AI Futures Project’s timelines update in response to METR: https://blog.aifutures.org/p/q35-timelines-update-looks-like-the

Data center-related ballot measures, 2026 (Ballotpedia): https://ballotpedia.org/Data_center-related_ballot_measures,_2026

Back to The AI Policy Podcast · Report a problem