Equity Automattic's 33-Hour coup, and can AI labs police themselves?
34 snips
Sep 18, 2026 Anthropic’s plan to slow frontier AI faces pushback from Nvidia and tough questions about independent oversight, market incentives, and political divides. Automattic’s chaotic 33-hour boardroom coup may leave Matt Mullenweg with more control over WordPress. Plus, May Mobility’s risky SPAC debut and DoorDash’s $425 million bet on automating food delivery.
AI Snips
Chapters
Transcript
Episode notes
AI Safety Consensus Meets Conflicted Incentives
- Independent evaluators, shared safety standards, and international coordination form the core of Dario Amodei’s proposal.
- Nvidia’s Jensen Huang pushed back while benefiting financially from continued AI acceleration, complicating his role as an industry mediator.
Why AI Pacing May Not Mean Slowing
- Political and commercial incentives align against an AI slowdown, especially as Nvidia and the Trump administration emphasize competition with China.
- “Pace” sounds safer than “slow down,” while the proposed measures may impose no actual speed limit.
Why AI Self-Regulation Resembles Boeing Oversight
- Market discipline may fail because enterprise customers cannot easily switch providers and heavily funded labs can absorb losses.
- Sean O’Kane compared weak oversight to Boeing, where self-regulation and insufficient monitoring preceded catastrophic failures.