Google DeepMind releases Genie 4 world model with audio and real-time control

Google DeepMind unveiled Genie 4, an upgraded version of its world model research project. The system generates interactive video environments with synchronized audio, more photorealistic visuals, and the ability to steer simulations in real time using text prompts mid-generation. The release is limited to a research preview for trusted testers, reflecting Google's cautious approach to rolling out early-stage generative models.