Make some stuff.
Particles, fields, energy, space. Keep the ingredients deliberately small.
A tiny, unofficial universe laboratory
I built a universe.
It has particles. It has energy. It has a few rules. That’s about it.
The interesting part is what happens after we stop helping.
Make the rules simple.
Press Run.
Don’t help.
Watch carefully.
01 · The whole idea
It is stupidly simple.
Start with very simple things, give them local rules, press Run, and see whether anything interesting appears that we did not explicitly design.
Give it rules.
Define how those ingredients may interact. Prefer local rules over global scripts.
Press Run.
Once the world begins, resist the urge to make the result more interesting.
Watch.
Measure what persists, combines, copies, competes, adapts, or surprises us.
02 · The point
The Kid didn’t put this here.
If we program a creature, we have built a creature. The harder and more interesting experiment is to program only the underlying world and see whether creature-like organization appears on its own.
×We built it.Interesting engineering.
✓It happened.Potentially interesting emergence.
03 · The discipline
How little can we get away with?
The temptation is to add cells, DNA, goals, organisms, species and evolution because those make a world look alive. We want to begin lower down the ladder.
organismsDNAbrainsspeciesgoalsparticlesenergyforcessimple interactions
The interesting things should make themselves.
04 · Meet the universes
Most universes will be boring.
Excellent. A negative result is still a result. We want a record of the dead ends, the explosions, the near-misses and the truly strange worlds.
Everything exploded.
Too much attraction. Very short universe.
Nothing happened.
Stable, reproducible, profoundly dull.
It won’t stop spinning.
Persistent rotational structures. Cause TBD.
Something weird happened.
Interesting enough to measure before naming.
Universe #1842: “They appear to be eating each other.”
05 · What are we looking for?
A field guide to increasingly interesting things.
These are not claims or promised milestones. They are simply useful rungs on the ladder from “dots moving around” toward richer organization.
Structures
Patterns that persist long enough to become objects worth studying.
Chemistry
Objects combining into stable arrangements with new behavior.
Metabolism
Structures using energy or resources to maintain themselves.
Replication
Structures producing descendants without a hard-coded “copy” command.
Evolution
Heritable variation plus selection creating cumulative change.
Intelligence?
Far beyond the present experiment. Still too interesting not to put on the wall.
06 · From the inside
Same universe. Two very different perspectives.
A parameter casually chosen by a creator could look like a deep law of nature to an observer who exists entirely inside the world.
Why are the laws what they are?
Our simulated inhabitants would inherit constants and rules too. To them, those rules would simply be reality.
Does complexity need a designer?
Or can surprisingly rich global behavior arise from very small local interactions?
What counts as alive?
Before using the word, we should be explicit about which properties we actually observed.
We do not expect the simulations to answer these questions. We hope they make the questions better.
07 · Repeat, a lot
One universe tells you almost nothing.
Change one rule. Change one constant. Change one starting condition. Run it again. Then compare what changed and what stubbornly stayed the same.
#1021#1022#1023#1024#1025…
The point is not to find the prettiest universe. It is to understand which kinds of rules repeatedly create which kinds of behavior.
08 · Help build it
The basement has room.
KIB can become more interesting if people contribute different kinds of attention: observations, ideas, code, analysis, visualizations, criticism and eventually spare compute.
Watch
Explore runs and flag structures or behavior that seem genuinely unusual.
Think
Suggest experiments, simpler rules, better measurements or reasons our interpretation may be wrong.
Build
Contribute simulation code, analysis, visualization, tooling or infrastructure.
Investigate
Take one interesting run apart and try to explain what actually happened.
Donate compute
Later, idle machines could help explore more of the rule-space.
“I think you’re fooling yourselves, and here’s how to test it.”
09 · Suggest an experiment
Leave the Kid a note.
A good suggestion does not need to be complicated. “What if particles aged?” is enough. We can worry about the experiment design afterward.
10 · Basement rules
A few rules for ourselves.
Don’t fake emergence.
If we put a behavior in, say so. “Emergent” should not be a marketing word.
Don’t call it alive because it wiggles.
Interesting behavior and life are not synonyms.
Weird is not automatically important.
Most anomalies will be bugs, artifacts or ordinary statistics.
Negative results count.
A universe where nothing useful happens still constrains the search.
Show our work.
Rules, seeds, parameters, code and observations should be inspectable.
Prefer explanation over awe.
The goal is not “wow.” The goal is understanding why the wow happened.
11 · Come downstairs
Come explore with us.
Many universes. A ridiculous number of possibilities. One stubborn question: what can simple rules build when nobody tells them what to become?
The Kid made a universe.
The universe made something interesting.
The interesting thing started wondering where it came from.