Autoři

Sofiia Mikhailova, Olena Mikhailova

Anotace

The prototype combines a Unity-based first-person environment with virtual-reality interactionand AI-generated narrative feedback. Trigger zones initiate story events, while dedicatedscripts record decisions and coordinate reflection generation. The player’s choices andscenario context are converted into structured input for a language model, whose response isdisplayed within the experience. In the Q&A, the team characterised the machine-learningarrangement as unsupervised but did not name a model or dataset. A browser or web-pageroute was also discussed for remote demonstration, although the presenters said that itrequires a private key.

licence

Creative Commons BY-SA 4.0