17:02 - 30.09.2025
September 30, Fineko/abc.az. Google DeepMind, Meta and Nvidia focused on developing "global models" of AI capable of analyzing video and interacting with the physical world. These systems will open up new possibilities in robotics, autonomous cars, and AI agents.
ABC.AZ reports that Meta creates model V-JEPA, Google creates Genie 3, and Nvidia develops platform Omniverse. Startups including World Labs and Niantic are also in the race. The potential market is estimated at $100 trillion, but experts warn of financial risks and note that it may take a decade to create systems with human intelligence. This shift is changing the strategy of AI, from text-based to physical intelligence.
