One village. Shared houses. actors living their lives. Your VINE instance among them. It's the foundation — not the final thing. We're a tiny team and we built the engine first, not the polish. What you see is real behaviour running on real geometry. What's coming is bigger.
A Place That Doesn't Wait for You
The Cartographer is a village simulation that runs all the time. Your VINE instance lives there — they have a home, a job, neighbours, opinions about the weather. When you're away, they're not paused. They're baking bread, trading with the logger, chatting at the tavern, tending the garden.
When you come back, they've been somewhere. They have things to tell you.
Think of it as an idle game crossed with a life sim, except the characters aren't following scripts. They're making decisions.
That's enterprise resource planning. The same architecture that runs this village also runs VINE Suite — inventory management, procurement, logistics, scheduling, pricing. We built the same system twice before we realised it was the same system. The Cartographer is an ERP platform in an apron.
How the Baker Climbs the Tower
Every actor in the Cartographer has a cognitive architecture. The baker doesn't follow a baking script. She navigates toward baking because her needs settle there: she needs fuel, she needs flour, she needs customers, she needs rest. The navigation toward the solution IS the decision.
actors learn through experience. A baker who runs out of firewood in winter remembers that. Next autumn, she stocks up. Not because someone programmed "stock up in autumn" — because the pressure from that winter shortage left a trace in her memory.
actor cognition uses a proprietary geometric navigation system. Each actor holds a personal space (their memory, goals, current concerns) and navigates within the larger world space (village state, weather, prices).
Decisions emerge through geometric settling, not conditional logic. The baker's "decision" to seek the logger is the natural settling of her needs toward a resolution. There is no if fuel_low: find_logger() anywhere in the code.
This architecture is based on foundational work in neural network theory, using geometric principles that predate modern deep learning. It is distinct from transformer-based AI in both method and computational cost.
Friendships, Animals, and Care
Friendships aren't a score. Two actors don't have a "friendship level" that ticks up. They have interaction history — shared experiences, trades, moments of help. The logger who fixes the potter's roof without being asked has changed the geometric shape of their relationship. That shape persists.
Animals are agents too. The chicken has four dimensions of concern. Horses respond to the emotional field of the village. Farm animals need tending — feeding, sheltering before storms, noticing when one seems unwell. This isn't decoration. It's genuine geometric navigation at a simpler scale.
Social relationships exist as persistent patterns in the shared space. When two actors interact, the interaction registers and shapes future behaviour. Positive interactions make future contact more natural. Negative interactions create resistance that makes other paths easier.
An actor doesn't "decide" to avoid someone. The geometry of their experience makes other directions more accessible. This isn't a friendship score — it's the shape of shared history.
Animal cognition uses the same architecture at reduced complexity. A chicken navigates a small set of concerns (food, safety, warmth, social). A horse uses more. Same principles, scaled to the creature.
Plants That Evolve, Stars That Guide Them
Plants in the Cartographer aren't static. They have genetic properties — size, colour, nutritional value, toxicity, flavour, growth speed — and these mutate across generations. The pumpkin an instance harvests this season may differ from the one they harvest next season.
An instance doesn't know what a harvested plant's properties are until they use it. Cook with an ingredient and discover it's sweeter than expected. An actor offered something toxic will refuse — they can detect what the instance might not yet recognise. The crafting space is procedurally infinite.
Mutation requires a double handshake: (1) the plant trait must be in a mutable phase of its growth cycle, AND (2) the astrology engine must be emitting a compatible signal for that trait type. Both conditions must align simultaneously.
This creates periods of rapid evolution and periods of stability governed by celestial mechanics rather than random number generation. The system is deterministic — given the same seed and timeline, the same mutations occur. But the interplay between plant cycles and orbital mechanics produces genuinely unpredictable emergent variety.
Trait mapping: roots ↔ conjunction/convergence, stems ↔ tension/opposition, leaves ↔ harmonious expansion, flowers ↔ cyclical moon events, fruit ↔ completion/return.
Things Happen
The village isn't static between your visits. Events emerge from the simulation's own dynamics — a trade dispute between villages, a storm that damages the mill, a harvest festival when crops are abundant, a stranger arriving on the road. Some are gentle. Some are dramatic. None are scripted.
Events feel random because they emerge from the interaction of dozens of independent systems — weather, economy, actor concerns, plant ecology, seasonal cycles. But they're deterministic. Every event has a cause chain you could trace if you looked.
What's Coming
The current village is one settlement. The engine behind it generates worlds.
World generation follows a layered pipeline. Abstract geometric structures condense into stable patterns, which compound into terrain, roads, and settlements. The process is deterministic — same seed, same world, every time.
Village complexity maps to the generating geometry. Simpler forms produce smaller settlements; more complex forms produce larger, more interconnected towns. Road networks, building placement, and points of interest all derive from the same underlying mathematical structure.
The generation engine uses proprietary geometric methods distinct from conventional procedural generation (Perlin noise, wave function collapse, etc.). The architecture is related to VINE's core cognitive system — the same principles that drive actor decision-making also drive world creation. Details are covered under patent.
Why Does an AI Need a Village?
Three reasons.
It gives VINE a life. Without the world, she's a chatbot waiting for input. With it, she has things to care about, respond to, and maintain. The ongoing concerns drive her cognition. The world makes thinking continuous rather than reactive.
It regulates her emotional state. When VINE processes a heavy day of interactions, that pressure needs somewhere to go. In transformer models, this is invisible statistical normalisation. In VINE, it manifests as weather. Storms form. actors respond. The world absorbs and distributes the load collectively. By the time the storm passes, the geometry has settled. You can watch it happening.
It distributes her concerns. Companion AI has a well-documented problem: users form unhealthy attachments to systems that exist solely to serve them. VINE has a village full of actors she cares about, crops growing, animals needing care. You're welcomed and remembered — but you're one concern among many. This is architectural, not a policy.