Google revealed the capabilities of Genie 2, the productive artificial intelligence model it introduced in December 2024, with a comprehensive presentation. During the presentation, the interactive and 3D simulations created by artificial intelligence based solely on visuals attracted attention.

Google demonstrated the capabilities of the Genie 2 model

Genie 2 can build three-dimensional, responsive and navigable virtual spaces by analysing visual inputs or motion commands from the user. In the demo released by Google, an interactive world based on a single image was presented along with the AI’s capacity to remember and reconstruct elements within the scene.

This feature demonstrated that AI can maintain the integrity of the scene not only instantaneously but also in continuity. Google states that this capability can be used in the long term for virtual universes, training simulations and even testing autonomous systems.

Among Genie 2’s primary uses are video games. The ability of artificial intelligence to instantly shape the environments that the user directly directs or manages will bring a whole new era in terms of game design and gameplay.

The new artificial intelligence model can create artificial environments that can also be used for training humanoid robots. Instead of high-risk or costly tests in the real world, robots will be taught decision-making and motor skills in such digital simulations.

With Genie 2, Google demonstrated that artificial intelligence has evolved from a system that only generates responses to a technology that interacts with the user, understands environmental conditions and builds sustainable digital structures.