Google's Gemini Moves Beyond Text, Adding Interactive 3D to AI Responses

Google is pushing its Gemini AI toward a more visual and manipulable future. A new capability allows the chatbot to build interactive 3D models and simulations directly from user prompts. This transforms a query from a static answer into a dynamic object you can control.

For example, asking Gemini to illustrate the Moon's orbit around Earth generates a rotatable 3D model. Users can then adjust the orbital speed with a slider, hide the orbital path, or pause the simulation entirely. It represents a shift from receiving information to experimenting with it.

The update signals a broader industry trend where leading AI models are rapidly acquiring visual reasoning skills. In recent weeks, Anthropic enabled its Claude AI to produce charts and diagrams, while OpenAI's ChatGPT gained tools to visualize scientific concepts. Google's move, however, introduces a layer of direct user manipulation previously unavailable in Gemini, which was limited to interactive images.

To use the feature, Gemini app users must select the 'Pro' model in the prompt bar. Posing a question like 'show me a double pendulum' or 'visualize the Doppler effect' will prompt a 'Show me the visualization' button to appear, launching the interactive model. This development suggests a future where business analysis, training, and prototyping could be conducted through conversational prompts paired with malleable digital prototypes.

Source: The Verge