// the body · wiring

the wire

a model of human wiring that carves itself by use. sensory in, interneurons, motor out. send a signal — the wires that carry it strengthen, the rest fade. repeat it and a path appears. the wire is the trace of what ran. a toy that shows the rule, not a real brain.

signals 0 · strong wires 0 · unwired

blue = sensory in · green = interneurons · red = motor out. click an input to fire just that one. send the same signal a few times and watch the path carve.

// the body, by the numbers

~86 billion
neurons (an estimate, ~62–99B)
~10¹⁴
synapses (≈100 trillion)
~150 m/s
fastest signal, myelinated
~176,000 km
of myelinated fiber
20–40 nm
the gap — they never touch

// what you're seeing — and the science under it

  • the axon is the wire; myelin is the insulation. the fatty sheath lets the signal jump gap to gap (saltatory conduction) — faster, farther. (Purves et al., Neuroscience)
  • the synapse is a junction, not a weld. a 20–40 nm gap separates the two cells; the signal crosses it chemically. the wire is re-negotiable. (Cajal's neuron doctrine; Loewi 1921)
  • use lays the wire. Hebb (1949): when one cell repeatedly takes part in firing another, the link between them strengthens. that's the rule the model runs. (Hebb 1949, p.62)
  • "cells that fire together wire together" is a later paraphrase (Shatz 1992; Löwel & Singer 1992) — it drops Hebb's order: A fires before B. (Shatz 1992)
  • strength is physical. brief heavy use leaves a lasting jump in a synapse's efficiency — long-term potentiation. (Bliss & Lømo 1973)
  • the brain over-builds, then prunes. synapses peak in infancy ~50% above adult levels, then unused ones are cut. (Huttenlocher 1979)
  • the survivor is the trace. from a wide initial repertoire, use stabilizes some connections and removes the rest — selection, not a blueprint. (Changeux & Danchin 1976; Edelman 1987)

// what this is not

  • this is a schematic — it shows the principle (wiring by use), not a real brain. real wiring is chemical, vastly larger, and far more connected.
  • no complete synapse-level map of a human brain exists, or is yet feasible. the largest map to date covers ~1 mm³ of mouse cortex. (MICrONS 2025)
  • real connectomes are imaged from actual tissue, not generated by a rule. completed ones so far: a worm (~300 neurons), a fly (~140,000). (White 1986; FlyWire 2024)
  • the Human Connectome Project maps coarse tracts between regions, not synapses. the numbers above are estimates, not exact counts.

// sources

Azevedo & Herculano-Houzel 2009 · Goriely 2025 · Purves, Neuroscience (2nd ed.) · Marner et al. 2003 · Hebb 1949 · Shatz 1992 · Löwel & Singer 1992 · Bliss & Lømo 1973 · Huttenlocher 1979 · Changeux & Danchin 1976 · Edelman 1987 · Sporns, Tononi & Kötter 2005 · White et al. 1986 · Cook et al. 2019 · FlyWire / Dorkenwald et al. 2024 · MICrONS 2025.