DeepMind unveils Genie 4 world model with real-time physics simulation and native audio for robot training
CyberRobooo · x · 2026-10-28
Google DeepMind announced Genie 4, a new-generation world model with unprecedented realism and improved real-time interactive simulation, now in research preview for trusted testers before rolling out to Google AI Ultra subscribers (Project Genie) and Gemini API/AI Studio early-access partners.
Key points:
- Natively multimodal: generates music-like audio, ambient noise, and interactive dialogue alongside the simulated world — all sounds in demos are produced directly by the model
- Built on the Gemini family: Gemini Omni anchors simulations in real-world appearance and behavior, while other Gemini models provide reasoning to keep simulations coherent and continuous
- Dynamic simulation: set an event in motion and Genie 4 simulates what happens next — how the world responds, sounds, and how people within it move, react, and speak
- Robot training potential: demos include Apptronik's Apollo humanoid playing checkers and a UR5 arm pipetting liquid into a lab plate; one person at a computer could train and test edge cases for an entire robot fleet
DeepMind frames world understanding as crucial to AGI. The open question remains sim2real transfer — whether simulated performance holds up in real-world tasks.
More from Embodied
- Robots Coming for Your Jobs: A Playful Reaction to a Demo — alysha_lobo · 2026-10-28
- New benchmark tests whether home robots navigate more cautiously around babies than trash bins — SeoulNatlUniv · 2026-10-28
- Early adopter revisits AI hardware: still meh, runs hot without opening apps — alexmacgregor__ · 2026-10-28
- RobotAgentBench debuts to test open-world robot autonomy: manipulation ≠ autonomy — HennyJieCC · 2026-10-28
- Would you ship a policy trained mostly in generated worlds? AIFlow polls robot people — AIFlow_ML · 2026-10-28
- Re:Build CEO: robots attack a small slice of production cost, lean beats them — jaimalik · 2026-10-28