The Algorithms the Universe Runs On

Take a sheet of graph paper and a single rule. A square is alive or dead. Look at its eight neighbours: a live square with two or three live neighbours survives; a dead square with exactly three live neighbours comes to life; everything else dies, of loneliness or of crowding. That is the entire law—no hidden clause, nothing about shape or speed or strategy. And yet from that one rule, run on a blank grid, things begin to happen that nobody put there. Small clusters stabilise. Others blink forever. And some—the gliders—detach and walk across the plane, the same five cells reassembling one step over and one step down, a moving body made of nothing but the relentless application of a four-line law. John Conway called it the Game of Life, and it is not really a game. It is a demonstration that complexity is cheap, and that simplicity, iterated, is the most generative force there is.

This is the thing the twentieth century kept rediscovering in different rooms. The richness around us—the fern, the coastline, the spiral arm of a galaxy, the branching tree of an artery—does not require a rich set of instructions. It requires a poor one, applied without mercy and without end. The universe, on this reading, is less a finished painting than a running program: a few short rules and an enormous amount of time. And there is a second wonder folded inside the first. When you tally the numbers that govern these processes—the ratios, the angles, the exponents—the same few keep appearing in places that share no cause. The angle at which a sunflower packs its seeds is the angle at which a fir sets its needles. The law of how a river network branches is the law of how the airways branch inside your chest. Something is forcing the issue.

How Little It Takes

Conway’s grid was a parlour version of a deeper claim. In the early 1980s Stephen Wolfram began studying the simplest computers imaginable: a row of cells, each black or white, each updated by a rule that looks only at itself and its two neighbours. There are just two hundred and fifty-six such rules, and most are dull—they die out, or freeze, or repeat. But one of them, the rule he numbered 30, does something that still seems impossible the first time you watch it. Started from a single black cell, it generates a clean triangular order on one side and, on the other, a stream so thoroughly without pattern that it passes statistical tests for randomness. No randomness was added; the rule is fixed and every step is forced. Wolfram spent two decades on this and published it as A New Kind of Science, arguing that the right model for nature is not the smooth equation but the simple program, the rule iterated. Some of his grander claims remain unproven. The central one is hard to unsee: complexity does not need a complex cause.

The same lesson hides inside a question that sounds like a joke. How long is the coast of Britain? Benoit Mandelbrot took it seriously, and the answer is that it depends on the length of your ruler. A long rod cuts across every bay; a metre stick traces inlets the rod ignored; a centimetre wraps around pebbles. The coastline grows the closer you look, without limit, because the same roughness repeats at every scale—bays within bays, the part resembling the whole. Mandelbrot gave this a name, fractal, and a measure: a dimension that need not be a whole number. A coastline is not a one-dimensional line but something like 1.25-dimensional, crinkled enough to begin filling the plane. Once you have the idea you find it everywhere nature had to fit a great deal of surface into a little space.

Complexity does not need a complex cause. A short rule, run long enough, can outrun any shortcut you might use to predict it.

Give the rule a little grammar and it grows plants. An L-system is a handful of substitution rules—replace this symbol with that string, again and again—and read as turtle steps and turns, those strings draw ferns and trees barely distinguishable from the living ones, because a living bud is itself a rule for making two buds. Alan Turing, in the last paper he published before his death, asked how a featureless ball of identical cells could ever decide to grow a pattern. His answer: two chemicals, one activating and one inhibiting, diffusing and reacting at different speeds, will spontaneously break their own symmetry into spots and stripes. Reaction and diffusion, two ordinary processes, lay down the leopard’s rosettes and the zebra’s bands before the animal is born. The coat is not drawn. It precipitates, out of chemistry obeying a rule.

The Edge Where Order Breaks

Not every simple rule settles down. Some, pushed hard enough, fall apart in a way that turns out to be exquisitely organised—and the cleanest place to watch it is a single line of arithmetic. Picture a population that grows in proportion both to how many there are and to how much room is left. Take this year’s fraction of the maximum, multiply by a growth rate, multiply by what remains, and feed the answer back in as next year’s fraction. That is the logistic map. For a small growth rate it does the sensible thing and settles to a steady number. Turn the rate up and the sensible behaviour shatters by stages: first the population alternates between two values, a fat year and a lean one; then two become four, then eight, the period doubling and doubling faster, the windows of calm shrinking geometrically. At a precise threshold the doublings have piled up infinitely and the system tips into chaos—a sequence that never repeats and depends so sensitively on where it began that two starts a millionth apart diverge completely within a few dozen steps. Deterministic to the last decimal, unpredictable in practice forever.

Here the story takes its uncanny turn. In the mid-1970s the physicist Mitchell Feigenbaum measured how fast those doublings arrive—the ratio of the gap between one splitting and the next. He got a number: 4.6692, and on. Then he tried it on an entirely different equation, a different curve with nothing in common but the act of folding back on itself, and got the same number. The constant does not belong to the population model. It belongs to the road to chaos itself, the way pi belongs to every circle rather than to any one. A dripping tap, a heart sliding into fibrillation, a fluid the instant before it churns—vastly different systems approach chaos by the same staircase, set to the same universal beat. The behaviour is wild. The route in is law.

The Numbers That Will Not Leave

Which brings us to the constants, the handful of numbers that keep walking into rooms they were never invited to. Pi is the friendliest and the most quietly shocking. Of course it lives in every circle—but it also sits at the centre of the Gaussian, the bell curve that governs measurement error and human height and the drift of pollen on water, a formula with no circle in sight. Pi appears wherever there is rotational symmetry to respect, and probability summed over enough independent nudges has a rotational symmetry hidden in it. Then there is e, the number near 2.718 that is simply what growth looks like when it never pauses. A sum compounded every instant, a population breeding continuously, a hot cup cooling—anything whose rate of change is proportional to its size carries e in its bones, because e is the one base at which a quantity’s growth equals the quantity itself. It is the natural unit of becoming. And it keeps the most beautiful company in mathematics: push e to an imaginary power and rotate it halfway round and you get the identity that joins the five great constants in one breath, e^(i*pi) + 1 = 0. Growth, rotation, the imaginary, unity, and nothing, locked into a single true sentence.

“The miracle of the appropriateness of the language of mathematics for the formulation of the laws of physics is a wonderful gift which we neither understand nor deserve.”—Eugene Wigner, 1960

And then there is phi, the golden ratio near 1.618, where the recurrence stops being a curiosity and starts looking like a law of life. Watch a plant decide where to put its next leaf. If it set each one a simple fraction of a turn from the last, the leaves would line up in spokes and shade one another. So instead the growing tip rotates by the golden angle, about 137.5 degrees—the circle divided in the proportion of phi. Why that angle, of all the infinite choices? Because phi is, in a precise sense, the most irrational number there is: the hardest of all to approximate with any fraction. An angle built from it never repeats. No leaf is shadowed by the one below; no seed wastes the gap beside it. Pack seeds at that angle and they spiral outward in the interlocking arcs of a sunflower head—the figure above—their counts running through the Fibonacci numbers, because Fibonacci is just phi learning to count. The pinecone does it. The pineapple does it. Your stem of basil does it. Optimal packing has one answer, and the plants all found it.

Phi is the most irrational number there is. An angle built from it never repeats—so no leaf is ever shadowed by the one below.

The deepest recurrence is the branching. Hold up a river delta, a bolt of lightning, the bare winter canopy of an oak, the bronchial tree cast from a lung, the web of vessels feeding your retina. Strip the labels and you cannot tell them apart. They share not just a look but a law—the same fractional dimension, the same power relating how the branches thin as they divide. And there is a reason. Each is solving one problem: how to reach a whole volume from a single source with the least possible length of channel. Nature ran that optimisation a thousand separate times, in water and air and blood and fire, and arrived a thousand times at the same exponent. Not because the lung consulted the river. Because there is only one good answer, and persistence finds it.

Discovered, or Demanded?

So why? Why should the same short rules and the same few numbers reach across scales that share no cause—from the spiral of a galaxy to the spiral of a seed head? In 1960 the physicist Eugene Wigner gave the puzzle its enduring name, the unreasonable effectiveness of mathematics in the natural sciences, and confessed he could not explain it: mathematics worked out by pure thought kept turning out to be exactly the mathematics the world ran on. One old answer is Plato’s—the patterns are real, more real than the things that wear them, and we do not invent mathematics but remember it, the way you recover something you always somehow knew. Against it stands the formalist, for whom mathematics is a human game of marks and rules, no more discovered than chess. Discovered or invented: the fault line has not moved in two and a half thousand years, because each side explains exactly what the other cannot.

The boldest modern answers refuse the gap entirely. John Wheeler coined the phrase it from bit—the proposal that information is the true ground floor, every particle drawing its existence from binary answers to yes-or-no questions. Wolfram presses the program metaphor to the limit: the cosmos is not described by computation, it is computation, a rule grinding forward step after step. And Max Tegmark goes furthest, holding that physical reality does not merely obey mathematical structure but is one—that in the end there is no difference between the equation and the thing. If he is right, Wigner’s mystery dissolves, because asking why mathematics fits the universe becomes like asking why the universe fits itself. Here is the honest difficulty, and it cuts the other way: none of these grand pictures has paid its way in a single new prediction. They are stances, beautiful and unfalsified and so far idle. We are pattern-finding animals who evolved to see the regular and ignore the rest, and it is entirely possible the effectiveness is partly a selection effect—that we notice the corners of nature that yield to our equations and look past the vast turbulent remainder that does not. The map fits the territory, a skeptic might say, because we drew the map by tracing the parts we could already see.

What Persistence Obeys

But there is a reading that keeps the wonder while letting some of the mystery go, and it starts by noticing what these constants have in common. Phi is not a magic number stamped on creation; it is the answer to a question—the best way to pack things that keep arriving. Pi is not a cosmic signature; it is what you get the instant anything turns. e is not written into the stars; it is what constant-rate growth cannot help becoming. The branching exponent is what least-cost distribution always converges to. Each is a fixed point: the place a whole class of processes ends up no matter where it starts, the way water finds the one low spot in a valley. They recur not because something repeats them but because anything that grows, packs, branches, or optimises is solving a problem with only one solution—and that solution is the number we then find waiting there. So the universe is perhaps not written in mathematics, like a text in a language. It may be closer to say that mathematics is the name we give to the regularities any persistent structure is forced to obey. A thing that lasts must be stable; to be stable is to sit at a fixed point; and the fixed points have addresses, and the addresses are numbers.

The constants are not the cosmos speaking in code. They are the small set of places it is even possible to stand.

And that turns the real marvel inward, to something we walk past every day. The miracle was never that the equations are elegant, or that pi keeps turning up uninvited. The miracle is that there is anything stable at all—that in a universe entitled to be formless and brief, structures arise that hold their shape long enough to grow, to branch, to pack a seed head, to fold a coastline, to read a sentence about themselves. The same few rules and the same few numbers recur because they are the narrow grooves through which persistence has to pass, and persistence is the rarest and most astonishing thing there is. You are made of branching that found the cheapest path, of spirals that wasted no room, of chemistry that learned to keep itself going against the long pull toward nothing. The wonder is not that the universe runs on a handful of algorithms. The wonder is that it kept running—and that some of what it produced woke up, looked around, and recognised the rules.