DEV Community

Cover image for What if a virtual world wrote its own story?
Donovan Chartrain
Donovan Chartrain

Posted on Originally published at donovan-dev-web.github.io AI-assisted

What if a virtual world wrote its own story?

From The Sims to RimWorld, the best game stories were never written. What if we built a whole world around that idea?

Every player has one of those stories. A colony that falls apart because one stubborn cook refused to leave the kitchen. A village that thrives because two neighbours happened to share a habit. You know the kind: the story nobody wrote, and everybody retells.

Those stories are not scripts. They are what happens when systems meet.

Stories from systems

The Sims made a simple idea popular: give characters needs (hunger, tiredness, a wish for company) and let them cope. In RimWorld, colonists have moods, traits and relationships, and an in-game storyteller adds events, but the drama comes from how those colonists react to each other. Dwarf Fortress built a reputation on the same ground: players tell tales of fortresses whose fate nobody planned.

What these games share is a quiet promise: if you design the rules well, the world will surprise you.

A lab, not a game

LIVEX takes that promise seriously, and then changes the goal. It is not a game. It is an open-source laboratory for watching what happens when agents with needs, memories and beliefs share a world.

The recipe is short. You set the rules. You start the simulation. You watch. An agent gets hungry; it remembers where food used to be; it believes a neighbour might have some. Nothing says what happens next. A gathering, a habit, an imbalance: whatever appears comes from local rules colliding.

Capture d'écran de l'écran d'accueil du Launcher LIVEX.

Researchers call this emergence: group-level behaviour that you cannot find in any individual taken alone. An ant colony has no boss and still organises its paths. A flock of birds draws shapes nobody draws. LIVEX is an attempt to make that kind of phenomenon visible, and then explainable.

Watching is not understanding

Here is the part I find most exciting. Seeing a pattern is easy; knowing why it appeared is the real work. So LIVEX separates the world from the tools that observe it. One component runs the simulation; another, called ECHOS, reads what happened and helps trace it: who interacted with whom, what came before, which chain of events led to this.

Capture d'écran du Launcher LIVEX lors du lancement d'une expérience.

It is the difference between enjoying a match and understanding it: the passes, the gaps, the sequence that produced the goal.

Honest about the magic

None of this makes the inhabitants alive in any human sense. A behaviour that looks like intention is not intention. LIVEX is not a model of human society, not an artificial general intelligence, and not a finished game. It is a deliberately abstract world, built to explore mechanisms.

Why dream about it

Because a world that can surprise its creator is a world that can teach something. If you can watch collective behaviour appear, then trace it back to its causes, curiosity becomes a method. And because the project is open, anyone can look under the hood.

If you would like to see the project from the inside, start here:
https://donovan-dev-web.github.io/LIVEX/en/ .

The code and documentation live on GitHub:
https://github.com/donovan-dev-web/LIVEX .

Donovan Chartrain | Creator and developer of LIVEX

Top comments (0)