On Thursday, Scale AI and the Center for AI Safety (CAIS) released Humanity's Last Exam (HLE), a new academic benchmark aiming to "test the limits of AI knowledge at the frontiers of human ...