DeepMind's Genie 4 world model adds audio and more realistic people to simulations

Key facts

• Genie 4 generates photorealistic and interactive worlds with real-time audio.
• Users can change scene parameters like weather or add elements mid-simulation.
• The model maintains consistent scenes for up to two minutes.
• Gemini 3.8 Flash acts as a director, planning steps and checking frames for Genie 4.
• Potential uses include gaming, interactive learning, robotics, and film prototyping.

Google DeepMind has released Genie 4, an updated world model offering enhanced photorealism, interactive capabilities, and real-time audio generation. Integrated with Gemini 3.8 Flash, the model allows users to modify simulations dynamically, such as changing weather or adding objects, and maintains scene consistency for up to two minutes. Potential applications span gaming, education, robotics, and film prototyping.