On May 19, at the I/O 2026 developer conference, Google DeepMind announced it will officially integrate Google Street View data into its Project Genie general-purpose world model. This integration aims to create highly immersive, interactive virtual environments where users can experience real streets while simulating weather changes or observing street scenes across different seasons. The core breakthrough of Genie lies in spatial continuity. The model not only generates environments based on real-world street views but also supports free viewpoint switching, expanding observation perspectives beyond vehicle-mounted cameras to pedestrians or robots. For instance, newly deployed robots in London could familiarize themselves with light refraction effects through simulation beforehand; autonomous driving company Waymo can leverage this system to train responses to extreme scenarios. Leveraging over 28 billion street view images collected by Google across 110 countries over two decades, this platform is gradually becoming a new infrastructure for robot training, game-based education, and autonomous driving simulations. Currently, this feature has been rolled out first to U.S.-based Google AI Ultra subscribers, with global access expected within weeks. Researchers from DeepMind acknowledged that the project remains experimental: generated visuals resemble video games rather than photo-realistic imagery, physical laws are yet fully embedded, and logical flaws such as clipping still occur during simulations. The team anticipates significant improvements in model accuracy within six to twelve months. As algorithms evolve, the combination of Street View and Genie holds promise to reshape digital world construction methodologies.