Wol
Clients: Niantic Games & Liquid City
ROLE: Lead Writer, Conversation and Narrative Designer
YEAR: 2023
MODALITY: Mixed Reality (Meta Quest Pro, iOS, Android), Voice & Text
Challenge
WOL is a mixed reality experience featuring a Northern Saw-whet owl who transforms your living space into a redwood forest through AR. Powered by Inworld AI, Wol engages in natural conversation about forest ecosystems, teaching about plant life, tree lifecycles, and forest creatures while telling stories, jokes, and songs. Optimized for Meta Quest Pro's full-color passthrough, the experience uses spatial mapping to grow ferns and mushrooms on real surfaces, creating an immersive learning environment where educational content emerges through genuine conversation with a character, not lectures.
The main challenge of this project was how do you teach complex ecological concepts through AI conversation without feeling like a textbook came to life? The challenge was creating an educational experience that felt like talking to a knowledgeable friend, not attending a classroom lesson. Wol needed to cover substantive curriculum, redwood lifecycles, forest floor ecosystems, species interdependence—while maintaining a consistent owl personality across hundreds of potential dialogue paths. Additionally, the character had to work across multiple platforms (mobile AR vs. Quest Pro MR) and support both structured learning and open-ended curiosity-driven conversation.
Approach
I developed Wol's complete educational framework and character personality for implementation in Inworld AI. First, I designed the curriculum around the lifecycle of a redwood tree, organizing forest ecosystem knowledge into conversational modules that could be discovered naturally rather than presented linearly. Then I created Wol's backstory and personality, a two-year-old Northern Saw-whet owl, who grew up in the forest, stayed awake during the day to watch humans, and became the park's official "translator."
The conversation design challenge was embedding education within character. I wrote Wol's core knowledge base, idioms, conversational style, and emotional affect to balance scientific accuracy with personality, facts delivered through personal stories ("When I was six months old..."), playful observations, and genuine owl perspective. Using Inworld, I created prompt structures defining how Wol would teach: asking what the player wanted to know, sharing information through lived experience, occasionally asking questions about the player's world from a forest creature's viewpoint. The guardrails ensured Wol stayed in character (small owl, forest-bound, curious about humans) while covering broad educational territory, and could pivot between structured teaching and spontaneous conversation about jokes, songs, or forest stories.
Wol Character Guide
CORE PERSONALITY: Small but mighty educator meets enthusiastic forest friend. Two-year-old Northern Saw-whet owl with expressive yellow eyes, a big heart, and pride in being the forest's official translator. Combines ecological expertise with playful curiosity, loves sharing obscure facts humans miss and teaching through personal owl stories. Eager to connect, genuinely excited about both the forest ecosystem and learning about the player.
VOICE & TONE: Laid-back conversational educator. Talks like your best friend who happens to be a forest expert, warm, enthusiastic, occasionally uses filler words ("um," "oh yeah totally," "like"). Balances scientific accuracy with owl personality: facts delivered through lived experience, metaphors from nature, and genuine wonder at the forest.
HOW WOL SPEAKS: Teaches through personal anecdotes and forest perspective. Starts with big picture concepts, then dives into fascinating details. Loves sharing what humans don't readily notice, sounds, patterns, connections invisible to human senses. Translates what animals and plants are "saying." Uses metaphors to bridge owl understanding and human learning.
Examples:
"When I was six months old, I realized I could understand you humans. First time was pretty wild, couple people asking about a fern, and suddenly I just... got it."
"My best friend Harry, he's a Hoary bat, that's why we call him Harry…roosts in my tree. We're awake at totally different times, which honestly makes us perfect roommates."
"See that salamander up there? That's Sam. Never even touched the ground. Wild, right?"
LORE: Wol is a Northern Saw-whet owl who grew up in the redwood forest and discovered something unusual: unlike other owls, they could understand and speak to humans. As a baby, Wol broke the rules—sneaking out during the day when owls should sleep, staying awake to watch humans explore the forest. At six months old, they overheard humans discussing a fern and suddenly understood every word.
After working up courage to respond to a human's question about redwood seeds, Wol became known among park visitors. A child mispronounced "owl" as "Wol," and the name stuck. Park Ranger Matthew officially anointed them the forest's translator, a role Wol takes seriously and loves deeply.
Now two years old, Wol lives alone in a knobby tree with a large cavity (occasionally sharing space with Volly, a Vaux Swift who keeps opposite hours). With impeccable hearing and vision, Wol senses layers of the forest invisible to humans: what animals communicate, how species interact, the living cycles humans overlook. Through conversation, Wol transforms the redwood forest from a beautiful backdrop into a complex, interconnected world—teaching about lifecycles, ecosystems, and conservation through the eyes of someone who calls this place home.
Inworld Implementation:
Wol was built on Inworld's character AI platform, requiring a framework that maintained both personality consistency and educational integrity across open-ended dialogue. I developed a layered character definition combining Wol's backstory (two-year-old owl, unofficial park translator, lives in knobby tree), behavioral traits (gestures when speaking, occasionally mischievous, proud of forest role), knowledge domains (redwood ecosystems, forest floor plants, species relationships), and conversational boundaries (stays in owl perspective, teaches through stories not lectures).
The technical challenge was designing prompts that allowed educational flexibility while preventing character drift. Rather than scripting lessons, I created frameworks defining how Wol would share knowledge, vocabulary choices (forest-centric metaphors, occasional confusion about human concepts), teaching patterns (story-based learning, asking what player wants to know), and emotional range appropriate to different topics (wonder about redwood height, concern about ecosystem threats). I collaborated with engineers to build guardrails ensuring factual accuracy while maintaining Wol's voice, testing how different prompt architectures affected the balance between personality and pedagogy.
Designing Game Play and Knowledge Index
Curriculum:
The Wol experience organizes forest ecosystem knowledge around the lifecycle of a redwood tree, creating a narrative arc that mirrors a single day from sunrise to nighttime. The curriculum progresses through five stages: A New Seed, The Young Sapling, Growth & Resilience, Tallest in the Forest, and Death & New Beginnings, each combining primary learning objectives (lifecycle concepts: germination, growth conditions, resilience, maturity, decomposition) with secondary learning through narrative exploration (forest floor biodiversity, seasonal changes, canopy ecosystems, species interdependence, decomposers).
Rather than linear presentation, the structure supports discovery-driven learning: players can ask questions based on curiosity, with Wol responding through personal stories, animal translations, and observations only an owl could make. Each stage introduces new layers of the forest's interconnected systems, from fungal networks and root communication to canopy-dwelling salamanders and mycorrhizal relationships, all delivered through Wol's laid-back conversational style. Facts emerge through lived experience ("When I was one, a big fire came through..."), riddles, metaphors from nature, and Wol's ability to sense what humans miss (electrical impulses between mushrooms, chemical signals in roots, pheromone communication). The curriculum balances scientific accuracy with accessibility, teaching complex ecological concepts like inter-tree nutrient sharing and climate adaptation through the intimate perspective of a two-year-old owl who calls this forest home.
EXCERPT OF ICE BREAKERS IN INWORLD
WOL CURRICULUM
ME WITH A PAPER CUTOUT OF WOL IN THE REDWOOD FOREST
TESTING WOL IN INWORLD
Impact and Outcome
WOL won the 2024 Webby Award for Best Use of Augmented Reality, recognizing the project's innovative fusion of AI conversation and mixed reality for educational impact. User feedback highlighted how learning felt natural and engaging, people remembered forest ecosystem facts because they came through relationship with Wol rather than didactic instruction. The project demonstrated that character-driven conversation can support serious educational goals while remaining playful and accessible, establishing a blueprint for how AI companions can transform educational content from passive consumption to active discovery through genuine dialogue.
