Google’s AI tool, Genie 2, is a “large-scale base world model” capable of generating “an infinite variety of playable, action-controllable 3D environments” from a single prompt. ‘picture.
Genie 2 can create different perspectives, such as a first-person view, isometric views or third-person driving videos, as well as “complex 3D visual scenes” with interactive objects like doors and explosive barrels .
Physical effects including smoke, gravity, lighting and reflections can also be “quickly” prototyped and played by a human or “AI agent” using the keyboard and mouse. According to a report detailing the advanced technology, this allows artists and designers to quickly create prototypes, “which can begin the creative process of designing the environment, thereby speeding up research.”
“With Genie 2’s out-of-distribution generalization capabilities, concept art and designs can be transformed into fully interactive environments,” the report explains. “This allows artists and designers to quickly create prototypes, which can accelerate the creative process of designing the environment and further accelerate research.
“While this research is still in its early stages and there is substantial room for improvement in agent and environment generation capabilities, we believe that Genie 2 is the way forward to solve a problem “structural framework for training embodied agents safely while achieving the scale and generality required to progress toward AGI.”
The full report, including examples, is available at Google Deepmind subsite.
Earlier today, British specialist media publisher Future has signed a strategic partnership with OpenAI to use its ChatGPT tool in its sales, marketing and editorial activities.