Google DeepMind unveils Genie 4 interactive world model with native audio

What happened.Google DeepMind announced Genie 4, a multimodal world model capable of simulating interactive environments with synchronized audio and real-time natural language steering.

Why it matters.Real-time world generation bridges game physics simulation and autonomous robotic training. Providing synchronized acoustic and visual physics allows robotics teams to train spatial navigation policies in synthetic environments without manual audio engineering.