Google DeepMind Introudces Genie 2, an AI That Generates Playable 3D Worlds

DeepMind, Google's AI research division, has introduced Genie 2, a groundbreaking model capable of generating an "endless" range of interactive 3D worlds.

As the successor to DeepMind’s earlier model, Genie, released earlier this year, Genie 2 can create real-time, playable environments based on a single image and text description, such as "A cute humanoid robot in the woods." This innovation places it in the same category as other emerging models, including those from Fei-Fei Li’s World Labs and the Israeli startup Decart.

DeepMind boasts that Genie 2 can generate a "vast diversity of rich 3D worlds," where users can engage in actions like jumping and swimming using a keyboard or mouse. The model is trained on videos, enabling it to simulate object interactions, animations, lighting, physics, reflections, and even NPC behavior.