A computer was trained to play Qbert and immediately broke the game in a way no human ever has

It found a bizarre way to rack up 1 million points by playing the game "in what seems to be a random manner" and making the entire stage flash.
9to5Mac

Apple publishes Human-Centered Machine Learning workshop videos

Today, Apple published on its Machine Learning Research blog, select recordings from its 2024 Workshop on Human-Centered Machine Learning (HCML), highlighting its work on responsible AI development.