I subscribe to the position taken by 406 scientists about what is needed to ensure AI can’t cause harm to humans and our institutions.
METR evaluates frontier AI models to help companies and wider society understand AI capabilities and what risks they pose.