AIPolitics

Jeremie Harris on what happens after a real AI disaster, and why China still won’t trust us | Crosspost from NonZero

Read full article at source ↗

Based on the episode description, not a full transcript. Source evidence retrieved 05/11/2026, 19:03:24 UTC.

Briefings

Jeremie Harris, CEO of Gladstone AI, argues that AI safety requires an 'AI Chernobyl' event where the president fears for their own life, triggering a policy shift. Harris claims Washington has not yet experienced this, citing his six-year bet that AI scaling would not bend. He criticizes Open Philanthropy for discouraging government briefing due to fear of 'powerful weapon' disclosure, while crediting it for improving the problem's visibility. His assessment of OpenAI suggests 'evaporative cooling' of safety-minded staff, and he questions Sam Altman's sincerity regarding warnings. The episode covers government response, past mistakes, and China, but lacks specific data on China's AI capabilities or US policy timelines.

Key points

• Jeremie Harris argues that Washington has not yet experienced an 'AI Chernobyl' disaster severe enough to flip the official stance on AGI risk.
• Harris claims Open Philanthropy discouraged outsiders from briefing Washington due to fear officials would hear 'powerful weapon' and launch a Manhattan Project for superintelligence.

Context

This podcast episode features Jeremie Harris, CEO of Gladstone AI, discussing AI governance challenges. The source material is an unpaid cross-post from NonZero, originally published on the NonZero website. Harris is a coauthor of a 2024 State Department-commissioned action plan on AI risk and has spent six years betting that AI's scaling curve will not bend.

Limitations

Evidence is limited to a podcast transcript without specific data on China's AI capabilities or US policy timelines. No additional background on the AI safety ecosystem is supplied beyond Harris's claims about Open Philanthropy's role and government briefing challenges.

Generated by Bittrees-hosted qwen35-2b sha256:aaf42c8b7c3c on 05/11/2026, 19:59:24 UTC. Based on publisher-provided feed evidence; may not cover the full original article. Check the source for details.

View archived briefing on IPFS
bafybeihq6sv7tljfw3e2w2qib2ywj3muijsrlryedzmoesezfqqwkwubwy