Testing Out Google Genie 3D Model
In this single‑speaker walkthrough, the presenter demonstrates Google Genie’s ability to generate an explorable 3D video‑game‑style world from a single starting image. The speaker uploads a fantasy‑themed portrait created in Gemini, then provides textual prompts describing the environment (colorful fantasy world with stream, grass, trees, rocks, ancient columns) and the character (white male avatar with brown hair, beard, blue plaid shirt, jeans, sword on back). After clicking “create world,” the system instantly renders a navigable scene where the avatar walks toward the camera, the camera can be rotated, and the character can jump with the space bar. The demo runs for 60 seconds, showing that each frame is generated in real time as the camera moves, with the system remembering the overall layout. The presenter notes visual limitations (pixelated graphics) and playful physics quirks, such as being able to walk on water as if it were land. Overall, the video showcases the current capabilities and constraints of Google Genie’s real‑time 3D world synthesis.
