What Is It?

This model simulates how distributed observers with different interests collectively construct narratives around real-world phenomena in the Hollywood Hills / UCLA / Griffith Park area.

Inspired by realtime.earth (Guerin) and Narrative DNA (Bergquist / Corto / USC ETC).

How It Works

Observers (circles) wander the landscape carrying 14D interest vectors. Phenomena (pulsing diamonds) spawn at plausible locations. When K distinct observer types within range report on the same phenomenon, a narrative (pie-segmented hexagon) emerges. Colors show which observer types contributed; the number shows how many distinct types.

Narratives attract further observers, persist longer than phenomena, and reshape interests over time.

Phenomena Types

Mountain lion — Griffith Park, Wildlife+Crime+Pets
Murmuration — Lake Hollywood, Birding+Sunsets+Clouds
Wildfire plume — Hills ridgeline, Wildfire+Weather+Clouds
Rare orchid — Temescal Gateway, Wildflowers+Fungi+Urban Eco
Flash flood — LA River channel, Flood+Weather+Urban Eco
Meteor shower — Griffith Observatory, Night Sky+Clouds
Lost dog — Runyon Canyon, Lost Pets+Crime+Urban Eco
Plane flyover — LAX flight path, Planes+Clouds+Night Sky
Fungal flush — Franklin Canyon, Fungi+Wildflowers+Wildlife
Sunset spectacle — Mulholland overlook, Sunsets+Clouds+Weather

Things to Notice

Coverage gaps — areas with no observers where phenomena fade unseen.

Spatial clustering — observers congregate near active narratives, leaving other areas blind.

Observation range matters — too small and observers miss nearby events; too large and geography stops mattering.

Order Parameters

Richness — mean distinct observer-types per narrative

Cascade — fraction of observers engaged with ≥1 narrative

Echo — within-type / total sharing ratio

Coverage — fraction of phenomena generating ≥1 narrative

Half-life — mean age of active narratives

Diversity — Shannon entropy of observer interests

Live Stats

Model: Stephen Guerin (realtime.earth) × Yves Bergquist (Corto / USC ETC)
Built with Claude Code, Jan 2026

Richness
Cascade
Echo
Coverage
Half-life
Diversity