What is Emergent Life?

A plain-English introduction. No background in biology, computing or artificial intelligence is assumed.

The short version

Emergent Life is a continuously running artificial world. Small creatures live inside it, controlled not by a script but by a tiny neural network each one carries. They sense a little of what is around them, decide what to do, act, and either survive that decision or do not. Over many generations, the population's behaviour changes: some traits help an organism eat, avoid harm or reproduce more successfully than others, and those traits become more common. That process is evolution, and it is the only mechanism shaping what these creatures do.

What "emergence" means here

A handful of simple, generic rules govern how organisms sense the world, spend energy, move, eat, reproduce, mutate and interact with material around them. Nobody writes a rule that says "cooperate" or "explore" or "build". Whatever patterns of behaviour actually appear are a consequence of those generic rules interacting with each other and with a changing environment, not a feature anyone typed in directly. That is what "emergent" means: a property of the whole system that was not explicitly programmed into any one part of it.

Why this is different from a conventional game

In most games, a designer decides what should happen and writes code to make it happen: a creature "attacks" because someone programmed an attack behaviour. Here, nothing decides in advance what an organism ought to do. Each one runs its own small neural network against whatever situation it is in, and that network's connections were shaped by generations of selection, not authored by a person. There is no win condition and no designed story. There is a set of physical rules and a long history of what has happened under them.

Organisms operate under generic rules

The same rules for energy, movement, perception, manipulation and reproduction apply to every organism, on both islands, for the whole life of the universe. Nothing is hand-tuned to make a particular story happen. If two populations never meet, that is because nothing in this run caused them to, not because meeting was disallowed or scripted away.

Persistent, not restartable on a whim

Public Universe 1 is not reset between visits. Its state, history and identity persist across server restarts. What you see today carries forward from what happened yesterday, all the way back to the moment it began.

Why nobody knows exactly what will happen

The rules are deterministic given a starting seed, but the space of possible outcomes over a long run is far too large to predict by reasoning about the rules alone. That is true of a great many complex systems, biological evolution included. The honest position, and the one this project takes throughout its methodology, is to describe what has actually happened rather than what was hoped for or expected.

Go deeper: how it works →