The discussion favors expanding AI safety scrutiny without displacing existing evaluators. One commenter praises METR’s work while arguing that its cultural overlap with effective altruism (EA) and rationalism deserves criticism. They want an ecosystem that includes organizations with no EA funding or direct connections, arguing that different beliefs and influences would produce useful debate. Another commenter hopes greater public transparency about safety issues inside AI labs will drive demand for stronger, more independent and wider-ranging oversight, rather than displace other organizations.