The path to slow AI

Anthropic CEO Dario Amodei wrote a new blog post, that. Lays out 3 strategies to slow AI’s roll.

(1) involves “embedded evaluators” from third-party organizations like METR — evaluators who can verify that AI companies are actually following their pacing and safety commitments.”

(2) “leading AI companies “within democratic countries” to coordinate “common safety standards as well as limits on the rate of unchecked AI progress.”

(3) calls for “global coordination,” where the United States and its allies “attempt to coordinate with authoritarian governments, to the extent this is possible.” 

Tech Crunch