Google has recently taken significant strides in the realm of artificial intelligence with groundbreaking projects like Gemini 2.5, Veo 2, and Agentspace. In December, the tech giant introduced Genie 2, an AI that has garnered attention for its capability to create interactive 3D worlds. This development marks a pivotal step towards the creation of simulations and games fully crafted by AI.
Following the initial announcement that captivated both technology enthusiasts and the gaming industry, Google has now offered a comprehensive introduction to Genie 2. Demis Hassabis, CEO of Google DeepMind, appeared on the renowned US program, 60 Minutes, to demonstrate the impressive capabilities of this promising AI.
Genie 2, a generative diffusion model, can craft 3D environments that dynamically respond to user commands and movements within the virtual realm. It can also extrapolate from a given image, visualizing and transforming it into a navigable interactive environment. The demonstration on 60 Minutes vividly showcased these abilities. Furthermore, Genie 2 retains memory of the simulated scene components outside the player’s view, accurately reconstructing them upon their return.
While Genie 2’s primary applications are in gaming, its potential extends far beyond. The AI holds the promise of revolutionizing various sectors. For example, 3D worlds generated by Genie 2 could serve as test environments, enabling humanoid robots to learn real-world interactions. Additionally, such generative AIs could play a crucial role in the development and customization of virtual universes, such as the Metaverse.
Yorum Yap