News Summary

Meta has introduced its latest AI model, V-JEPA 2, designed to help AI agents and robots understand and interact with their surroundings. The model is an extension of the original V-JEPA, trained on over a million hours of video to enable real-world understanding and prediction, such as how gravity affects objects. V-JEPA 2 is reportedly 30 times faster than Nvidia’s comparable model and aims to improve the functionality of machines like delivery robots and self-driving cars by allowing them to navigate and perform tasks in real-time. This advancement is expected to reduce the need for extensive robotic training data, marking a significant step forward in AI and robotics.