Automattic's 33-Hour Coup, and can AI labs police themselves?

Anthropic chief executive Dario Amodei has laid out a plan he calls pacing the frontier, a response to a doomsday warning issued a week earlier by a researcher at his own company, according to TechCrunch AI. The proposal centers on independent safety evaluators and coordination among AI labs based in democratic countries, an attempt to build a shared brake on the pace of frontier model development without waiting for government regulation to catch up.

TechCrunch reports the plan has already drawn some support within the industry, alongside pointed pushback from Nvidia chief executive Jensen Huang. The publication does not detail the full substance of Huang’s objection in the material reviewed here, only that it was notably critical.

The same TechCrunch AI dispatch also covers Automattic, the company behind WordPress.com and Tumblr, where a leadership dispute concluded within 33 hours, per the outlet’s framing of events. Details of the Automattic matter beyond that timeframe are not elaborated in the summary available.

Taken together, the two threads TechCrunch bundles into a single dispatch, corporate governance at a legacy web company and safety governance at a frontier AI lab, appear in the same podcast-linked writeup without an explicit editorial connection drawn between them beyond their shared publication date, September 18.