Google has opened Project Genie to Google AI Ultra subscribers. This experimental prototype allows users to create, explore, and modify their own interactive worlds.
Google has opened Project Genie to Google AI Ultra subscribers. This experimental prototype allows users to create, explore, and modify their own interactive worlds.
According to the developers, the Genie 3 model generates environments in real time as the user moves and interacts with the world.
«It models physics and interactions for dynamic worlds, and its breakthrough consistency allows you to simulate any real-world scenario — from robotics and animation and science fiction simulations to location exploration and historical events,» Google says.
Project Genie is a prototype web application based on Genie 3, Nano Banana Pro, and Gemini that allows users to experiment with a «world model.» They can create a character and world using text or image prompts, and specify the mode of exploration—from walking to riding, from flying to driving, and anything in between.
«For more precise control, we’ve integrated World Creation with Nano Banana Pro. This allows you to preview how your world will look and tweak the image to fine-tune your world before you dive into it. You can also define the perspective for your character—for example, first-person or third-person—giving you control over how you perceive the scene before you enter it,» the developers say.
Users can also «remix» their interactive worlds
and make new interpretations using prompts. The developers have provided the ability to visit other users' worlds, which are available in curated collections.
Recall that the Google DeepMind team introduced a new version of its artificial intelligence model Genie 3 last summer. Access to it was only available to a small group of scientists and content creators, but now Project Genie can be tried out by Google AI Ultra subscribers in the US; it will be available to users in other countries later.
