Dario Amodei and other AI leaders want to ‘Pace the Frontier’ but…how?

A week after one of Anthropic’s researchers set off fresh doom talk around AI, CEO Dario Amodei is trying to answer it with policy instead of panic. His idea is to “pace the frontier” of AI development, which is a neat phrase for a messy problem: how do you slow the fastest builders without pretending everyone will happily stop at the same time?

Amodei’s plan leans on independent safety evaluators and on coordination among AI labs in democratic countries. That’s the core of it. Not a ban, not a freeze, but a framework that depends on outside checks and a bit of trust between rivals. Some people in the industry seem open to it. Nvidia’s Jensen Huang, though, has already pushed back.

That tension is exactly what the Equity hosts dig into here: if companies are asked to slow down, what does “slow” even mean, and who decides when a lab has gone too far? The episode also branches out to Automattic’s boardroom coup and a couple of big deals from the week, but the AI debate is the one with the real bite. Because once the frontier starts talking about pacing itself, the argument is no longer about fear. It’s about power.