Initially, the aim was to find out how easy and useful it is to use ChatGPT to write Unity C# scripts and implement them directly. ChatGPT should randomly distribute a number of points within a defined space. The points are manually replaced in Unity by a prefab (any object). As soon as the camera approaches, the points interact with the camera.

As this worked without any issues, the next experiment tested the limits of the Meta Quest 2 hardware. Various methods were tried to keep the computational load for a single object as low as possible, in order to place as many objects as possible. The shape of the object, the lighting and the scripts also have a significant impact on the hardware’s performance.

Finally, experiments were carried out with the existing scenes, and various features such as optical illusions, sound and other reactions were tested.