Making healthier choices is hard. But it’s also something that AI is uniquely positioned to solve, write Sam Altman and Arianna Huffington.

Over the course of four sessions, my patient, Stacey, and I worked toward her goals of sleeping more, checking her email less and spending more quality time with her family. She successfully developed ...

Several frontier AI models show signs of scheming. Anti-scheming training reduced misbehavior in some models. Models know they're being tested, which complicates results. New joint safety testing from ...