Playmaker of Visions

Playmaker of Visions is a procedurally generated VR experience that uses AI to make each session feel personal. Four AI models work together to power conversations with a virtual Playmaker, generate its voice, and create images and objects in real time. The world shifts and responds to each participant's choices as they make them.

The project was developed around cutting-edge AI tools, including an image-to-3D model released in March 2024, one month before the experience launched. It was among the first XR projects to use this kind of model. On a Meta Quest 3, AR passthrough and real-time 3D scanning let AI-generated content sit inside your actual physical space, responding to the geometry of the room around you.

You place the headset on your head and look out across the room. Nothing is out of the ordinary until a rift tears open in the space in front of you. A portal widens and the Playmaker begins to speak.

You are asked about your vision for the future. A selection of word bubbles appears around you: Sustainable, Electric, Renewable, Car. You select a few to describe your vision and push them into the Playmaker's portal.

Thousands of images pour out and begin to position themselves around the room, sticking to furniture and coating the walls and floors.

The images pop one by one, turning into 3D objects you can grab and interact with. The objects that emerge are as varied as the visions that created them.

As the experience draws to a close, you are invited to take a seat. The physical space around you slowly gives way to a virtual counterpart and you begin to rise toward the ceiling, then out across the building, high above the city.

Finally, watching the Earth from space, the scale of your vision is put into perspective. From up here, the stakes are obvious.