Lower builds upper; upper steers lower
The Computos is one continuous fabric, but it folds into layers. Atoms make molecules; molecules make cells; cells make minds. Each layer is built from the computations beneath it — and in turn it constrains and directs them. These panels show that two-way traffic, and what happens when a computational process grows, falters, or splits in two. Click any panel to run it.
Reality is layered, but the layers are not separate things — each is a way the layer below organizes itself. Click a level to see both directions of traffic: how it is built upward from finer computation, and how it reaches downward to constrain what its parts may do.
Every level is the level below, organized. Pick one to read its two-way traffic.
Upward causation (blue): the parts, computing, give rise to the whole — quarks make protons, neurons make thought. Downward causation (amber): the whole, once formed, sets the boundary conditions its parts run inside — a living cell decides which of its molecules' reactions proceed. Neither direction is more real. They are one process, read from two ends.
No blueprint, no foreman. Order appears because each part follows simple local rules, and those rules, run together, settle into structure. This is how a layer of the stack assembles itself from the one below.
e.g. crystallising sugar, a snowflake, cooling rock
Scattered molecules, jostling at random, lock one by one onto a growing lattice — each finding the site that costs least energy. The ordered crystal is computed by nothing but the rule "settle where you fit best."
e.g. a soap micelle, a cell membrane, a lipid bilayer
Each molecule has a water-loving head and a water-fearing tail. That single preference, multiplied across thousands, drives them to fold into a sealed wall on their own — the first boundary a cell ever needs.
A computational process is not fixed. It can grow stronger and more productive; it can falter as errors accumulate; and it can stop computing altogether — at which point its effects do not vanish but pass into the system that outlives it. Choose a path and watch the same process take it.
Strengthen: practice, growth, production · Impair: error, damage, disease · Cease: the pattern halts
Strengthen: connections multiply and the network grows denser and faster — the computation gains capacity, the way a trained skill or a thriving organ does. Switch modes to see the other paths.
The deepest move in the lifecycle: a computation that copies its own state and splits into two. Replication is computation producing more computation — one process becoming a population.
e.g. a dividing cell, the DNA double helix, yeast budding
The cell copies its internal state — its genetic instructions — then pinches in two, handing each daughter a full set. What was one computing process is now two, each able to copy again. Run it and watch one become many.
Building, constraining, organizing, growing, faltering, dividing — these are not separate phenomena but the same computation, viewed across scales and across time. The stack is alive because the Computos never stops folding into new levels of itself.
