Top AI researchers like Fei-Fei Li and Yann LeCun are developing world models, which don't rely solely on language.
Android Bench ranks AI models based on their ability to complete real Android coding challenges.