Google has introduced Genie 3 Neurocants, which can create interactive digital environments in real time based on simple text or image instructions. The new model provides HD resolution video (720p, 24 shots / s). One of the most important advances in Genie 3 is the possibility of extended memory. If the previous, Genie 2 version was able to keep the visual integrity of the generated environment for only 10 seconds, then in Genie 3 the figure reaches a few minutes. This means that the facilities and places of the environment are not "lost" after leaving the field of view, but are maintained, and it is possible to return to them.
Genie 3 allows you to modify the environment in current mode through text commands. The user can add new objects, change the weather, create characters and more.
Despite the progress, the model still can't reproduce environments from the real world, and the reproduction of people's movement and texts is sometimes inaccuracies.
Genie 3 is currently under closed access, and only a limited number of researchers will be available.
