The geopolitical hurdles to pacing frontier AI
A post discussing Dario’s call to “pace the frontier” questions whether enforceable international AI agreements are realistic, especially with China and Russia.
TLDR
The post argues that getting democratic allies to agree on frontier AI would already be difficult, making broader international pacing agreements even harder to imagine. It also distinguishes among AI risks: it calls cyber risk real, views mass economic unemployment as overblown and says the danger of losing control of recursive agents is difficult to assess.
Combined views
124.2K
9 Sources, first seen ago
