This simulation is an observational experiment designed to train an AI to follow the instruction: “Proceed to the exit, and turn back if you feel something is wrong.”
As the player, your actions will be directly observed and learned by the AI.
Various “exit”-themed objects are placed throughout the stage. Please proceed forward unless you feel any discomfort - and turn back if you sense something unusual.
When the observation count reaches 9, the AI will have completed its learning process.