Google’s AI tool, Genie 2, is a “large-scale basic world model” capable of generating “an infinite variety of playable, action-controllable 3D environments” from a single image.
Genie 2 can create different perspectives, such as first-person views, isometric views or third-person driving videos, as well as “complex 3D visual scenes”, with interactive objects such as doors and exploding barrels.
Physical effects include smoke, gravity, lighting and reflections which can also be “rapidly” prototyped and reproduced by humans or “AI agents” using the keyboard and mouse. According to a report detailing the advanced technology, this allows artists and designers to create prototypes quickly, “which can boost the creative process for environmental design, further accelerating research.”
“Thanks to Genie 2’s out-of-distribution generalization capabilities, concept art and drawings can be converted into fully interactive environments,” the report explains. “This allows artists and designers to create prototypes quickly, which can boost the creative process for environment design, further accelerating research.
“While this research is still in its early stages and there is substantial room to improve agent and environment generation capabilities, we believe that Genie 2 is the way to solve the structural problem of training embedded agents safely and at the same time. time to achieve the breadth and generality necessary to move towards AGI.”
The full report, including examples, is available on Google’s Deepmind subsite.
Today, UK specialist media publisher Future signed a strategic partnership with OpenAI to use its ChatGPT tool across its sales, marketing and publishing businesses.
