An embodiment of the invention relates to an artificial intelligence system comprising at least two servers used to generate instructions for a non-player character (NPC) to interact with a user depending on whether the user is within 1 metre of the NPC in the virtual world and configured to determine a plurality of interactions in response to the received input data related to an action, a facial expression, a gaze, a text, a speech, a gesture, an emotion or a movement detected from the user.
