IAPS @sevdeawesome says METR has more capacity to evaluate frontier AI models than the entire US civilian government: "Low regret things are just increasing visibility. Bring in CEOs and senior researchers to testify in Congress. What are your timelines, how capable are these models, what are you talking about internally." "The US just doesn't have that much state capacity to track and measure models. METR is a nonprofit, and they have the most watched benchmark that is most relevant for…