Jakarta, INTI – A new wave of artificial intelligence innovation has emerged from Google through its AI division, DeepMind. The company has introduced Genie 3, an AI model capable of building 3D virtual worlds similar to video games using nothing more than text instructions. This technology is expected to bring fresh experiences not only in entertainment but also in education and simulation-based training.
Instant Virtual Worlds Without 3D Assets
Genie 3 is an upgrade from its predecessor, Genie 2, designed to instantly create virtual environments. Users simply type a text description such as “a classroom with a blackboard and large windows” and the system will automatically generate a 3D world according to that prompt.
Unlike traditional 3D asset creation, which can be time-consuming, Genie 3 eliminates the need for manual design or separate asset downloads. Once the virtual world is created, users can immediately explore and interact with various objects inside it.
Enhanced Interaction and Object Memory
Unlike Genie 2, which limited virtual world exploration to just 10–20 seconds before resetting, Genie 3 allows for several minutes of interaction, delivering a more immersive experience that feels like truly being inside the environment.
One of Genie 3’s standout features is its ability to remember the position and details of objects for up to one minute, even when the user looks away. For example, writing on a blackboard will remain unchanged when the user looks back, instead of randomly altering.
This memory also applies to user interactions, such as moving objects or leaving marks in the environment. As long as the memory time limit has not been reached, all changes remain intact, enabling a more natural flow of interaction.
More Realistic Visuals and Dynamic Worlds
Genie 3 also brings significant visual improvements, supporting resolutions up to 720p at 24 frames per second, resulting in sharper worlds and smoother, more natural movements.
Additionally, Google has introduced a new feature called promptable world events, allowing users to modify elements of the virtual world through text commands. Weather conditions can be changed, new characters can be added, and specific interactions can be triggered instantly making the virtual world more dynamic and personalized.
Conclusion
The launch of Genie 3 by Google DeepMind marks a major step forward in AI technology that simplifies the creation of virtual worlds. With its ability to build interactive environments from text, advanced object memory, enhanced visuals, and customizable worlds, Genie 3 opens up new possibilities for entertainment, education, and interactive simulation training.
Read More:ElevenLabs Launches AI Feature to Create Music from Text for Global Creators