Google has launched an experimental prototype called Project Genie that allows users to create and explore interactive digital worlds generated by artificial intelligence.
This tool utilizes a recently developed world model known as Genie 3. The model is designed to simulate the environment, generating in real time the path a user follows as they move through a virtual space. Users can create a world by entering text prompts or uploading images, define their characters, and choose a method of locomotion, such as walking or flying.
The company said the project is part of its research into advanced artificial general intelligence (AGI), arguing that such systems need to learn how to navigate diverse environments.
The first release will be in limited quantities. Access is currently only being rolled out to users 18 and older in the United States who subscribe to Google’s top AI service, called Google AI Ultra. The company says it plans to expand to more regions in the future.
Google described Project Genie as an early research prototype with some known limitations. The company says the generated world may not be completely realistic, may not follow prompts or real-world physics exactly, and may be difficult to control in some cases. Currently, each world session created is limited to 60 seconds.
