What if artificial intelligence didn't begin in a Silicon Valley lab, or even with Alan Turing's 1950 paper — but with a bronze giant patrolling the shores of ancient Crete?
The idea of creating thinking, moving, autonomous beings is not a modern obsession. It is one of humanity's oldest dreams, encoded in myth long before the first circuit was ever etched. This is where the story of AI truly begins.
The Bronze Giant and the Clay Man
In Greek mythology, Talos was a massive automaton forged by Hephaestus — god of the forge — and set to guard the island of Crete. He circled the shores three times a day, hurling boulders at unwelcome ships. What makes Talos remarkable isn't just his existence in legend, but the detail with which the Greeks imagined his engineering: he was animated by ichor, divine fluid that flowed through a single vein running from his neck to his ankle, sealed by a bronze plug. Remove the plug, drain the fluid — and the machine dies.
That is not magic. That is a mechanical intuition about internal systems, circulatory design, and single points of failure. The ancient Greeks were, in their own language, describing an autonomous robot with a known vulnerability — a problem that remains central to AI safety research today.
Across the Mediterranean, Jewish mystical tradition produced the Golem: a creature of clay animated by inscribing the Hebrew word emet (truth) on its forehead. Erase one letter and it becomes met — dead. The Golem was built to protect, but in most tellings it grows too powerful, too literal in its obedience, and ultimately has to be destroyed by its own creator before it causes irreparable harm.
If that narrative arc — creation, utility, loss of control, destruction — sounds familiar, it should. It is the central anxiety of every serious AI safety conversation happening right now.
From China to the Islamic World
The dream was not confined to Europe. The ancient Chinese philosophical text Liezi describes a craftsman presenting King Mu of Zhou with a life-sized mechanical figure that could walk, sing, and wink. When the king ordered it disassembled, the craftsman revealed an interior of leather, wood, glue, and lacquer — artificial organs arranged to produce human behaviour. Remove the heart, and it could not speak. Remove the liver, and it could not see. The text raises a question that still has no clean answer: is the difference between a living being and a sufficiently complex machine a difference of kind, or merely of complexity?
In twelfth-century Turkey, the Islamic engineer Al-Jazari built what can only be described as the world's first programmable musical robots. His mechanical boat carried a drummer, two harpists, and a flautist who could perform different rhythmic patterns depending on the configuration of pegs on a rotating cylinder inside the drum. The program was encoded in the physical arrangement of components — a conceptual step that connects directly to the punched cards of Babbage's analytical engine and the stored programs of modern computers.
The Philosophers Who Mechanised the Mind
The seventeenth century brought the most consequential shift of all: philosophers began arguing not just that bodies were machines, but that minds might be too.
René Descartes declared the human body a perfect automaton — a machine built by God, operating entirely by physical principles. He drew the line at the soul, arguing that language and true reasoning could never be mechanised. But he acknowledged this was a practical limitation, not a principled one. He was, in effect, writing the terms of a debate that would not be settled for three more centuries.
Thomas Hobbes went further. In the opening pages of Leviathan (1651), he wrote simply: "reasoning is nothing but reckoning." Thought is computation. If that is true, and if computation can be performed by machines — as his contemporaries were already demonstrating — then a machine could, in principle, think. Leibniz built on this to dream of a characteristica universalis: a universal symbolic language in which all disputes could be settled not by argument but by calculation. His motto: "Let us calculate!" That vision is the direct intellectual ancestor of every logic gate in every processor ever made.
The Thread That Runs Through Everything
Hero of Alexandria built a programmable cart in the first century CE — its path predetermined by the arrangement of ropes around axles. Leonardo da Vinci designed a mechanical knight around 1495 that could sit, stand, and move its arms. Blaise Pascal built the first practical calculator in the 1640s and immediately felt uneasy: it produced outputs indistinguishable from a calculating human mind, yet had no understanding whatsoever of what it was doing. That gap between correct output and genuine understanding would later become John Searle's Chinese Room argument — one of the most debated thought experiments in the philosophy of AI.
What is striking, looking across this sweep of history, is not the diversity of these traditions but their unity. From Hephaestus to Hero, from the Kabbalistic rabbis of medieval Prague to the rationalist philosophers of seventeenth-century Europe, the same questions recur: Can intelligence be created? Can reason be mechanised? Can life be manufactured?
The engineers who built the first electronic computers in the 1940s were the inheritors of Al-Jazari's musical automata, Leibniz's calculating machine, and Pascal's uneasy Pascaline. The challenges we associate with AI today — alignment, control, consciousness, responsibility — have precedents stretching back thousands of years.
Artificial intelligence is not an accident of the late twentieth century. It is the continuation of humanity's oldest and most persistent dream.
The full article explores each of these traditions in depth — the complete mythology of Talos and the Golem, the engineering genius of Al-Jazari, the philosophical frameworks of Descartes, Hobbes, and Leibniz, and the mechanical wonders of the Renaissance — tracing the unbroken thread from ancient myth to the threshold of the modern computer age.
This is an excerpt. Read the full article (~3500 words) on ishistory.pages.dev →
Continue reading: AI Before Computers: Myths, Legends, and Mechanical Marvels
Top comments (0)