Let the Monster Speak

Let the Monster Speak is an interactive, AI-based educational game designed for families. It creates a space where parents and children can explore complex social issues together through the imaginative lens of monster narratives.
West
Digital Future
Service & Innovative Design
Intelligence
C1
Next Nature Museum
Noord Brabantlaan 1A
5652LA

Entrance fee

Location partially paid

Opening hours

11:00
-
18:00
11:00
-
18:00
11:00
-
18:00
11:00
-
18:00
11:00
-
18:00
11:00
-
18:00
11:00
-
18:00
11:00
-
18:00
11:00
-
18:00
Free Wifi Free wifi available
Toilets Toilets available
Wheelchair Friendly Fully wheelchair accessible
Wheelchair Friendly Toilet Wheelchair friendly toilet available

Follow

Monster Narratives

The narrative of this project is grounded in research on how monsters are created and why humans need monster narrative. It unfolds through four distinct storylines, each centered on a specific type of monster:

Scapegoat: monsters invented to cover up wrongdoing
Mirror: monsters used to define and reinforce what is considered “human” Mascot: monsters exploited for economic or symbolic value
Hero Maker: monsters that exist to produce heroes

Each storyline corresponds to critical social themes such as rumor-making, discrimination and hierarchy, media manipulation, and post-human perspectives.

In this exhibition, the Mascot narrative is featured, exploring how media manipulation operates within the commercialization of monsters. The aim is to make these abstract ideas approachable and discussable, especially for younger audiences. Visitors are invited to shape the story by letting the “monster” speak, revealing the hidden truths we often disguise through fantasy.

AI narrative authoring approach

The narrative is generated by a tailored Large Language Model, and the authoring process applies Orchid, an authoring approach developed by Zhen Wu and her research teammates, Serkan Kumyol, Shing Yin Wong, and Prof. Tristan Braud from the Division of Integrative Systems and Design at HKUST.

The authoring approach focuses on three aspects:

1.⁠ ⁠Story graph: The overall narrative structure, organizing the relationship between different narrative nodes and their sequences that guide the narrative's development.

2.⁠ ⁠In-nodes LLM generation rules for each narrative stage (node), including world settings, characters, player interactions, and narrator behaviors.

3.⁠ ⁠The conditions for switching between two adjacent nodes.

Scaffolded by the tool, these elements are integrated into the LLM-driven narrative algorithm, guiding it to generate dynamic narratives based on the player's input.

See related research publication: Orchid: A Creative Approach for Authoring LLM-Driven Interactive Narratives, in C&C '25: https://dl.acm.org/doi/10.1145/3698061.3726906

Experience Flow

In this experience, players first interact online with an AI narrative system that generates unique monster stories and corresponding AR monster images. After selecting a theme, they are guided through a co-creative storytelling process with the AI, reflecting on social dynamics, power structures, and media influence.

Afterward, players receive a physical package that extends their digital experience into the real world. The package includes a research booklet, a game board corresponding to the chosen storyline, character cards, a personalized storybook generated online, and 3D monster figures. These elements invite families to continue exploring and playing with their own monster narrative in a tangible and imaginative way.

Experience Flow
Experience Flow
Game Board
Game Board
AI Narrative System
AI Narrative System

Hosted by Xiaomin Fan and Zhen Wu

Xiaomin Fan is a cross-disciplinary designer with a background in Narrative Environments, focusing on storytelling and experience design. Her work explores how AI can be integrated into narrative-driven design to create interactive experiences. Zhen Wu(Yoyo) is a researcher, designer, and interactive artist interested in playful interaction, sensorial interfaces, and emergent art. She is a PhD candidate from HKUST. Her research focuses on AI ambiguity and sensing curation.

Colofon

Concept designer , Experience designer, Researcher
Xiaomin Fan
AI narrative and interaction program developer
Zhen Wu