Google DeepMind announces Genie 4, a real-time world model

Summary

Google DeepMind announced research on Genie 4, a new world model for real-time interactive simulation. Users can steer scenes with natural language, such as moving the camera, changing the weather, or introducing characters, while the model generates ambient sound, music, and dialogue on the fly. Gemini family models anchor the simulation in real-world appearance and keep it coherent. Access starts with a limited group of trusted testers, followed by Google AI Ultra subscribers and early partners in the Gemini API and AI Studio.

Why it matters for our work

World models that create safe virtual environments are foundational for training the next generation of robots and agents. The limitation is that today it is an early-stage rollout limited to trusted testers.

Translated from the Korean original. Summaries may be translated and edited. Commentary reflects our perspective; forecasts remain the source’s views.