The Mirror
The follow-up to the Large Labor Model: the same dataset — 480 occupations, taken apart into 4,811 tasks, scored against six machine capabilities — turned into a mirror. Type your job and read it task by task: the hours a machine could already fill, the hours it can only speed up, the hours it cannot yet enter. The task tables are open (CC BY 4.0), and corrections from people who have done the job feed the next version.