Google has introduced the Genie 2 AI tool, capable of creating a fully playable 3D environment from just one prompt image

Genie 2, an artificial intelligence model by Google, is described as a "large-scale foundation world model" that transforms a single image prompt into limitless, action-controllable 3D environments.

This tool can generate various perspectives such as first-person, isometric views, or third-person driving scenes, while also creating intricate 3D visuals with interactive features, including doors and explosive barrels.

Physics effects, such as smoke, gravity, lighting, and reflections, are easily prototyped and can be interacted with by either humans or AI agents using a keyboard and mouse. As per a report, these functionalities help artists and designers rapidly prototype, enhancing the creative process for environment design and speeding up research.

The report further notes that Genie 2, thanks to its generalization capabilities, allows for the transformation of concept art and sketches into fully functional environments, aiding swift prototyping and bolstering creative workflows in environment design.

Though still in its early phases with much to advance in agent and environment generation, there is belief in Genie 2's potential to address the structural challenges of training embodied agents safely while achieving the breadth necessary for progress towards AI development.

Comprehensive details and illustrations can be found in the full report, accessible via Google's Deepmind page.

In related news, Future, a UK-based media publisher, announced a strategic collaboration with OpenAI to deploy its ChatGPT technology across sales, marketing, and editorial divisions.