tl;dr: Google DeepMind announced Genie 4, a real-time interactive world model that generates photorealistic simulations with native audio, including ambient sound and interactive dialogue. It integrates with Gemini models (notably Gemini 3.8 Flash as a guiding agent) to enable multi-step reasoning, physics simulation, and natural-language scene editing, with applications in robotics training, game prototyping, and filmmaking pre-visualization. Currently released as a research preview to trusted testers and rolling out to Google AI Ultra subscribers, it still has limitations around long-term coherence, object permanence, and complex physical interactions.