The Weird Power of Simple Rules: What Cellular Automata Reveal About Reality

Imagine you’re in a classroom where there’s a strange new fashion trend. Some students wear hats, some don’t. After each class, everyone checks the two students sitting immediately to their left and right, and follows one simple rule: You’ll wear a hat next class if exactly one of your two neighbors is wearing one now. If both neighbors have hats, that’s too popular—you won’t. If neither has a hat, hats are out—you won’t.

One day, only the student in the middle shows up with a hat. What happens next?

If you watch class after class, something surprising emerges. The hats don’t just disappear or spread evenly. Instead, a repeating pattern ripples outward from the middle—triangles of hats and no-hats that grow and shrink in a regular dance. A single simple rule, followed by each student looking only at their immediate neighbors, produces a pattern that no one planned and that covers the whole row.

This isn’t just a classroom game. It’s an example of something called a cellular automaton (CA for short), and it raises questions that have fascinated philosophers, scientists, and mathematicians for decades.

What Is a Cellular Automaton?

A cellular automaton is really just a grid of simple cells—each one can be in one of a few states (like on/off, black/white, alive/dead). Each cell follows a simple rule that looks only at its nearest neighbors. All cells update at the same time, over and over. That’s it. That’s the whole machine.

The weird thing is what happens when you run it.

The classroom example is one of the simplest possible cellular automata: a one-dimensional line of cells (your classmates), two possible states (hat or no hat), and a rule that looks only at immediate neighbors. If you map it out with black squares for hats and white for no-hats, and stack the rows for each class period, you get a pattern that looks like a series of triangles sprouting from the middle.

Now here’s the puzzle that got people interested: How can something so simple produce something so complex?

A single rule, applied locally by each cell, generates a global pattern that would be hard to predict just from reading the rule. You’d have to actually run it to see what happens.

The Four Kinds of Behavior

In the 1980s, a physicist named Stephen Wolfram decided to systematically explore all possible rules for the simplest kind of cellular automaton (one-dimensional, two states, looking at one neighbor on each side). There are exactly 256 such rules. He ran them all and noticed something interesting: their behavior fell into four broad classes.

Class 1 rules produce boring uniformity. Everything quickly becomes the same—all black or all white—and stays that way.

Class 2 rules produce simple patterns that repeat. You get stripes or checkerboards that cycle through a few arrangements.

Class 3 rules produce chaos. The patterns look random and never settle down—like static on an old TV screen.

Class 4 rules are the most interesting. They produce patterns that are partly regular and partly chaotic. Structures form, move around, interact with each other, and sometimes disappear. One famous Class 4 rule, called Rule 110, was later proven to be so powerful that it can perform any calculation a computer can perform—it’s a universal computer hiding inside a simple grid of cells.

The fact that some of these simple systems can do universal computation (meaning they can simulate any computer program, if you set them up right) was a huge surprise.

The Game of Life

The most famous cellular automaton is two-dimensional. It’s called Conway’s Game of Life (or just Life), invented by mathematician John Conway in 1970.

Life takes place on an infinite grid of squares. Each square is either alive (black) or dead (white). The rule is simple:

A dead cell becomes alive if it has exactly three live neighbors.
A live cell stays alive if it has two or three live neighbors.
Otherwise, the cell dies (of loneliness or overcrowding).

That’s it. But from this tiny rule comes an entire universe.

If you start Life with a random scattering of live cells and let it run, you’ll see remarkable things happen. Stable structures called “blocks” form and sit still. “Blinkers” pulse back and forth between two shapes. “Gliders”—tiny clusters of five cells—travel diagonally across the grid forever. “Eaters” can swallow gliders that crash into them, then return to their original shape.

People have spent decades cataloging these patterns, calling it Life’s zoology. Conway himself speculated that if you ran Life on a large enough grid with random initial conditions, eventually intelligent, self-reproducing “organisms” might emerge. Whether that’s true is an open question—but the fact that it’s even a plausible speculation tells you something about how rich this simple system is.

Life has also been proven to be a universal computer. That means, with the right initial configuration, it can simulate any computer program. A tiny grid of cells following one rule can, in principle, run Minecraft or write poetry—it would just take a ridiculously long time.

Why Philosophers Care

Cellular automata raise deep questions that go beyond computer science.

Emergence

Here’s an unsettling thought: if you knew the rule of a cellular automaton but hadn’t seen it run, could you predict what patterns would appear? For many rules, the answer is no. You’d have to actually simulate it step by step—there’s no shortcut. This is called computational irreducibility: the only way to find out what happens is to let it happen.

This connects to a big philosophical question about emergence. Patterns like gliders in Life aren’t programmed into the system—they just appear. A glider is a pattern of five live cells that moves across the grid. But no single cell “knows” it’s part of a glider. Each cell just follows its local rule. Yet the glider is real in an important sense: you can predict its behavior, you can interact with it (eaters can swallow it), and the whole system would behave differently without it.

So are gliders “real”? Or are they just patterns we project onto the grid? Some philosophers argue that detecting these patterns is an objective, mathematical process—not just something we make up. Others say that whether something counts as a pattern depends on who’s looking.

Free Will and Determinism

Cellular automata are completely deterministic. Every cell’s next state is fixed by the rule and its neighbors’ current states. Nothing random. No choice. Yet if you watch Life run, you might describe what’s happening in terms of gliders “moving” and “avoiding” eaters. You might say the glider “changed course” to avoid destruction.

Does that mean the glider has free will? Probably not. But the philosopher Daniel Dennett uses this as an example to argue that even in a fully deterministic universe, we can meaningfully talk about things like choice and avoidance at a higher level. The glider really does avoid the eater—that’s a true description of what happens, even though at the bottom level everything is just cells following rules.

Dennett says this shows that determinism doesn’t imply inevitability. Even in a world where everything is determined, some things are avoidable—depending on which level you describe them at.

Is the Universe a Cellular Automaton?

Here’s where things get really speculative. Some scientists and philosophers have suggested that the universe itself might be a cellular automaton. The physicist Edward Fredkin proposed what he called the “Finite Nature Hypothesis”: that space, time, and everything else are fundamentally discrete and finite, and that the universe works like a giant cellular automaton updating itself step by step.

If that were true, then everything we see—quarks, trees, people, stars—would be patterns in this cosmic grid, like gliders in Life. The laws of physics would be the CA rule. And the fact that the universe is computationally irreducible would mean that the only way to predict the future is to actually live through it—no shortcuts, no secret formulas.

Most physicists think this is unlikely. But nobody has disproved it. The idea raises fascinating questions: If the universe is a computer program, what runs it? What counts as a “step”? And if the universe contains universal computers (like our brains, or actual computers), does that mean the universe can simulate itself? (That gets tricky.)

What’s Still Open

Cellular automata research is full of open questions. The Edge of Chaos hypothesis suggests that interesting complexity—like life—arises at the boundary between order and chaos. Some CA rules produce boring order, some produce boring chaos, but the really interesting ones (Class 4) live in between. This might tell us something about why life exists: perhaps living systems naturally evolve toward this edge.

But the hypothesis is controversial. The measure that was supposed to identify the edge (a number called λ) doesn’t work reliably. Some rules that should be on the edge turn out to be boring, and some that should be boring turn out to be complex. The relationship between simplicity and complexity is still poorly understood.

Another open question is whether the Principle of Computational Equivalence (also from Wolfram) is true. It claims that almost all processes in nature that aren’t obviously simple are computationally equivalent—they can all simulate each other, given enough time and space. That would mean a brain, a weather system, and a star are all, in a deep sense, doing the same kind of thing. But critics point out that just because two systems can simulate each other doesn’t mean they’re equivalent—the mapping between them might be so complicated that it misses the point.

Key Terms

Term	What it does in this debate
Cellular automaton	A grid of simple cells that update simultaneously based on local rules
Emergence	When large-scale patterns appear that aren’t explicitly programmed into the rules
Computational irreducibility	When the only way to predict a system’s behavior is to run it step by step
Universal computation	The ability to simulate any computer program (some CA have this)
Edge of chaos	The hypothesis that interesting complexity happens at the boundary between order and random chaos

Key People

John Conway – British mathematician who invented the Game of Life in 1970, showing how a simple 2D CA produces astonishing complexity.
Daniel Dennett – American philosopher who used Life’s gliders and eaters to argue that determinism doesn’t rule out meaningful talk about choice and avoidance.
Stephen Wolfram – Physicist and computer scientist who systematically explored one-dimensional CA, proposed four classes of behavior, and argued that CA fundamentally change how we should do science.
Edward Fredkin – Physicist who proposed the “Finite Nature Hypothesis” that the universe is a cellular automaton.

Things to Think About

If you could prove that the universe is a cellular automaton, would that change anything about how you live your life? Should it?
A glider in Life can be described either as “five cells in a certain arrangement” or as “a thing that moves.” Which description is more true? Can both be true at the same time?
If a CA produces patterns that look like living creatures but is just following a simple rule, what would it take for us to consider those patterns actually alive?
The Edge of Chaos hypothesis suggests life might necessarily arise at a boundary between order and chaos. Do you think there could be an “edge” in real life—a perfect balance of structure and freedom that makes interesting things possible?

Where This Shows Up

Video games like Minecraft’s redstone circuits and Factorio’s conveyor belts are essentially cellular automata.
Traffic flow models use CA to predict traffic jams from local driving rules.
Biology uses CA to model how patterns form on animal skins (zebra stripes, leopard spots).
Artificial life researchers use CA to study how simple rules might produce self-replicating or evolving digital organisms.
Your screen—every pixel is a cell that updates based on its neighbors, though the rules are much more complex than the ones described here.