Google CEO Sundar Pichai has shared his views on the newly announced Project Genie AI tool. This new experimental tool from Google DeepMind allows users to create and explore interactive digital worlds using AI, and he says he’s been testing it himself. In a post shared on the microblogging site X (formerly Twitter), Pichai wrote: “Project Genie is a prototype web app powered by Genie 3, Nano Banana Pro + Gemini that lets you create your own interactive worlds. I’ve been playing around with it a bit, and it’s…out of this world. Rolling out now for US Ultra subscribers.”The announcement introduces the latest experimental offering from Google’s AI research arm to end users as the company continues competing in the AI race. The company said this feedback will help guide future development in areas such as AI research, simulation, and generative media. Google said it plans to continue expanding access to Project Genie and improving the technology over time.
Meanwhile,Google DeepMind CEO Demis Hassabis also took to X and wrote, “Thrilled to launch Project Genie, an experimental prototype of the world’s most advanced world model. Create entire playable worlds to explore in real-time just from a simple text prompt – kind of mindblowing, really! Available to Ultra subs in the US for now – have fun exploring!”In an official statement, the company said: “Starting today, we’re rolling out access to Project Genie for Google AI Ultra subscribers in the U.S (18+). This experimental research prototype lets users create, explore and remix their own interactive worlds.”
What is Google’s Project Genie, how it works and other details
Project Genie is a research prototype built on Genie 3, a general-purpose world model that Google DeepMind first previewed in August. “Unlike explorable experiences in static 3D snapshots, Genie 3 generates the path ahead in real time as you move and interact with the world. It simulates physics and interactions for dynamic worlds. At the same time, its breakthrough consistency enables the simulation of any real-world scenario — from robotics and modelling animation and fiction, to exploring locations and historical settings,” the company noted.With Project Genie, Google DeepMind is offering early access to this technology through a web-based prototype. Starting this week, the tool is being rolled out to Google AI Ultra subscribers in the US aged 18 or older. The company said access will expand to more regions over time.Project Genie generates interactive environments in real time. As users move through a world, the system predicts and creates what comes next based on their actions. This differs from traditional 3D experiences, which are usually pre-built and static.The prototype is powered by Genie 3, along with Gemini and Nano Banana Pro, two other Google AI systems. Together, these models allow users to build, explore, and modify digital worlds rather than simply view them. Google DeepMind said this work is part of its broader effort to develop artificial general intelligence, or AGI, which would require AI systems that can operate across many different real-world situations.Project Genie focuses on three main features. The first is world creation, also described as “world sketching.” Users can use text prompts or upload images to create environments. Characters, environments, and modes of transportation, such as driving, flying, or walking can all be specified. Prior to entering the setting, users can also select viewing perspectives, such as first or third-person point of views.World exploration is the second feature. After a world is created, users are free to explore it. The system generates the environment in real time as users navigate, adjusting in response to movement and camera changes.The third feature is world remixing. Users can build on existing worlds by modifying prompts or exploring examples provided in a gallery. Finished experiences can be recorded and downloaded as videos.Google DeepMind reported that Project Genie remains an early research prototype and has several limitations. The worlds generated may not always align with the prompt, the physics may not be realistic, and the characters may be difficult to control. Each generated experience is limited to 60 seconds.
