It was the spring of 1900. A group of Greek sponge divers, blown off course by a fierce Mediterranean storm, dropped anchor near the small island of Antikythera — a barren, wind-battered speck of land sitting between Crete and the Greek mainland. They were looking for shelter. What they found instead would take another century for the world to begin to understand.

Beneath approximately 45 metres of water lay the wreck of an ancient Roman cargo ship, dating to somewhere between 70 and 60 BCE. The ship was carrying statues, glassware, jewellery, and coins. Divers recovered them over the following months in one of history’s earliest organised underwater salvage efforts. The finds were extraordinary — marble sculptures, bronze warriors, amphoras. The kind of haul that would make any museum curator weep with gratitude.

But one lump of corroded bronze, caked in sea sediment and unremarkable to the eye, sat in the Athens National Archaeological Museum for nearly two years without anyone paying it much attention. It looked like rock. Like debris. Like nothing.

In 1902, an archaeologist named Valerios Stais noticed something. A gear. A single, perfectly formed gear tooth peeking through the encrusted surface of what everyone had assumed was just more salvage. He looked closer. Then he called his colleagues. And then the trouble began — because what they were looking at was not supposed to exist.

The Device That Broke the Timeline

Let me be precise about what the Antikythera Mechanism actually is, because this matters. It is not a clock. It is not a sundial. It is not a compass. It is a hand-cranked, analogue computing device — an astronomical calculator — capable of predicting the positions of the sun and moon, tracking the cycles of lunar and solar eclipses, displaying the phases of the moon, and calculating the dates of the ancient Olympic and Panhellenic Games, all simultaneously, across multiple interconnected dials.

It contains at least 30 interlocking bronze gears — some researchers believe the original mechanism had as many as 37. The largest gear is roughly the size of a man’s palm. The smallest teeth on the smallest gears are less than half a millimetre wide. The angular accuracy of the gear ratios is so precise that modern analysis has found them to be correct to within a fraction of a degree. The entire assembly fits into a wooden box roughly the size of a shoebox.

It was built, to the best of our dating, sometime between 150 and 100 BCE. That is not a typo. One hundred and fifty years before the birth of Christ. More than five hundred years before the fall of the Western Roman Empire. Roughly two thousand years before you are reading this sentence.

“Nothing remotely like this level of mechanical sophistication appears again in the historical record for over a thousand years.”

The next comparable mechanical device in human history — a geared astronomical clock with similar complexity — does not appear until medieval Europe, sometime around the 13th or 14th century CE. That is a gap of roughly 1,400 years. Fourteen centuries of silence. And then we simply… pick up where the Greeks apparently left off, as though someone pressed pause on a civilisation and forgot to press play again.

I want you to sit with the strangeness of that for a moment. Because this is not a gap in artistic style. It is not a gap in literary tradition or philosophical thought. This is a gap in mechanical engineering. In precision manufacturing. In applied mathematical computation. The kind of gap that does not happen organically in a civilisation that is simply developing at a steady pace. The kind of gap that implies something was lost.

What Would It Take to Build This Today?

Here is where I invoke my engineering background, because I think it is important that we do not romanticise this. Let us talk about what it actually takes to produce a device like the Antikythera Mechanism — not in the abstract, but in cold, uncomfortable specificity.

To manufacture the smallest gears in the mechanism, you would need metal working tools capable of cutting to sub-millimetre tolerances. You would need a design that embeds, in gear-ratio form, precise knowledge of the Metonic cycle (the 19-year period after which lunar phases repeat on the same calendar dates), the Saros cycle (the 18-year, 11-day eclipse prediction cycle), the Callippic cycle (a 76-year refinement of the Metonic cycle), and the synodic periods of the moon and sun — all simultaneously, in a single integrated mechanical system.

The Saros cycle alone requires tracking 223 lunar months. The mechanism does this with a 223-tooth gear. Not approximately 223. Exactly 223. The designer knew the number, knew the cycle, and built the knowledge permanently into bronze.

In 2006, a team from University College London used modern computer tomography scanning to read the inscriptions inside the mechanism — instructions, essentially — and discovered that the device came with a user manual. Engraved on its own surfaces. For a device this complex. In 100 BCE.

Let me put this in contemporary terms. The equivalent of what the Antikythera Mechanism was doing in the ancient world would be a handheld mechanical device today that — with no batteries, no digital processing, and no software — could tell you the current positions of all the planets, predict the next ten solar and lunar eclipses to within hours, calculate the date of the next Olympics, and adjust for the irregular orbit of the moon. And fit in your jacket pocket. And come with an engraved instruction card.

No such purely mechanical device exists today. We use computers for that. And we have had the Industrial Revolution, the Renaissance, two centuries of precision engineering, and the accumulated scientific knowledge of every civilisation that came before us.

The ancient Greeks had bronze chisels and fire.

The Question That Nobody Wants to Answer

The standard academic response to the Antikythera Mechanism goes something like this: it was an exceptional achievement by an exceptional individual — possibly Archimedes or someone in his tradition — and its loss represents a tragic accident of history rather than evidence of a broader advanced civilisation. The mechanism is a one-off. A fluke of genius. A star that burned bright and was extinguished.

I have enormous respect for that position. It is measured, careful, and appropriately humble in its resistance to overclaiming. But I confess it does not fully satisfy me. And here is why.

The mechanism did not arrive from nowhere. It did not materialise in a vacuum. The astronomical knowledge embedded in its gears had to come from somewhere — from centuries of careful observation, from mathematical frameworks sophisticated enough to calculate eclipse cycles accurately, from a tradition of applied science that produced minds capable of translating abstract astronomical knowledge into physical mechanical form. Genius does not build in a void. It builds on what came before.

The mechanism is not the mystery. The mechanism is the evidence. The mystery is everything that must have existed for the mechanism to be possible.

And that tradition — that infrastructure of knowledge and craft — appears to have died. Completely. Without leaving a successor, a school, an apprentice lineage, or a single comparable artefact for fourteen centuries. Fourteen centuries during which empires rose and fell, during which the library of Alexandria burned, during which an enormous proportion of ancient knowledge was simply… gone.

This is the pattern that keeps appearing when you look honestly at ancient history. It appeared with the Pyramids of Giza. It appears with the precision stonework of Puma Punku in Bolivia. It appears with the Nazca Lines of Peru, the megalithic structures of Malta, the water management systems of Mohenjo-Daro. Isolated peaks of sophistication, separated by vast valleys of apparent technological regression, with no clear developmental arc connecting them to what came before.

Lost Knowledge, or Lost Civilisation?

There are broadly two schools of thought on what the Antikythera Mechanism tells us, and I think it is worth laying both out honestly, without the sneering that sometimes accompanies this conversation.

The first school holds that it represents the upper bound of what ancient Greek science achieved

— a brilliant peak that was lost due to the destruction of records, the collapse of trade networks, the disruptions of conquest, and the simple, brutal randomness of what survives and what does not. Under this view, the mechanism is a tragedy of preservation, not a mystery of origin. We lost the knowledge because history is violent and careless with fragile things.

The second school — and this is where it gets genuinely uncomfortable for mainstream scholarship — holds that the mechanism is evidence of a broader intellectual and technological tradition that we have systematically underestimated, and that this tradition may itself have inherited knowledge from civilisations even older that we have not yet properly accounted for. Not aliens. Not magic. But organised human knowledge, developed over thousands of years, that was lost in catastrophes we have only partially reconstructed.

I am not asking you to pick a side. I am asking you to notice that neither answer is comfortable. The first requires us to accept that humanity’s greatest mechanical mind produced a device with no successors, no documented apprentices, and no comparable artefacts — and that this gap of over a millennium is simply bad luck. The second requires us to accept that our reconstruction of ancient human capability is significantly incomplete.

Neither of these is a small concession.

What the Modern World Owes an Ancient Greek

Here is something that is rarely said plainly: the functions that the Antikythera Mechanism performed in 100 BCE were not reliably replicated by European science until the 17th century CE. The prediction of lunar eclipses to the accuracy the mechanism achieves was not bettered by European astronomy until the work of Johannes Kepler and his laws of planetary motion, published between 1609 and 1619.

That is a 1,700-year gap.

We sometimes speak of the ancient Greeks as the fathers of Western civilisation — of philosophy, of democracy, of geometry. We teach their ideas in universities. We name our children after their heroes. But we rarely grapple honestly with the implication of an artefact like the Antikythera Mechanism: that in at least one domain of applied science, they were not our ancestors. They were our peers. Or, depending on how you account for the simplicity of our modern mechanical watchmaking traditions relative to this device’s original purpose, perhaps something more.

The UCL Antikythera Research Team, which has spent years using advanced imaging and three-dimensional modelling to reconstruct the complete mechanism, published findings in 2021

suggesting the front face of the device was an elaborate planetary display — showing not just the sun and moon, but Mercury, Venus, Mars, Jupiter, and Saturn, all moving across a zodiac dial in their correct relative periods. A complete model of the solar system. Mechanical. Hand-cranked. Two thousand years ago.

When NASA’s Jet Propulsion Laboratory calculates the trajectory of a spacecraft to Neptune, it uses equations that account for the gravitational influence of every significant body in the solar system. The ancient Greek engineer who built the Antikythera Mechanism knew, with sufficient precision to cast it permanently in bronze, the orbital periods of every planet visible to the naked eye. They were doing, in their own way, the same thing. Just without the rocket.

What else did they know that we have not found yet? And what did they inherit from civilisations even older than themselves?

The Questions We Have Not Finished Asking

The Antikythera Mechanism was discovered in 1900. It took decades before anyone even took it seriously. It took until the 1970s before X-ray analysis revealed the true extent of its internal complexity. It took until the 2000s before modern imaging technology gave us enough resolution to read its inscriptions. It has taken over a century of scholarship, and we still do not fully understand it.

That is a humbling thought. One corroded lump of bronze, the size of a hardcover book, has occupied the attention of mathematicians, astronomers, classicists, mechanical engineers, and historians for more than a hundred years. And the picture it reveals — of an ancient world of extraordinary scientific capability, followed by an extraordinary silence — raises questions that our current historical framework does not have clean answers for.

I have been writing this series on mystery and the human frontier because I am deeply suspicious of the comfort of settled answers. The pyramids unsettled me. The Antikythera Mechanism unsettles me more. Not because I think the ancients had supernatural help — I find that explanation intellectually unsatisfying, a way of avoiding the harder question. The harder question is this: what does it mean for human civilisation that knowledge of this quality was possible two thousand years ago, and yet was so thoroughly erased that it took fourteen centuries to recover even a fraction of it?

Because if knowledge that sophisticated can be lost so completely, what does that say about the fragility of what we think we know today? What are the Antikythera Mechanisms of our own era — the innovations, the discoveries, the intellectual architectures — that will be unrecoverable to the civilisation that comes after us? And have we, perhaps, already lost things we do not yet know we are missing?

The Unfinished Inventory

The Antikythera shipwreck has not been fully excavated. Expeditions in 2012, 2015, and subsequent years have recovered additional artefacts — including human skeletal remains in 2016, which are being analysed for DNA evidence about who was on the ship. The wreck is still giving up its secrets.

Which raises, for me, the most quietly terrifying question of all: if the most sophisticated mechanical device of the ancient world was sitting in a box labelled ‘miscellaneous bronze debris’ for two years before anyone noticed what it was — how many other mechanisms, scrolls, instruments, and artefacts have we classified, stored, or discarded without understanding what we held?

The ancient Greeks had a word — aporia — for the state of genuine, productive puzzlement. The recognition that you are standing before a question you cannot answer, and that this inability is itself the beginning of real understanding. I think the Antikythera Mechanism puts us in aporia. It should.

Because the deeper you look at it, the more you realise that what was found at the bottom of the Aegean Sea in 1900 is not just a remarkable ancient artefact. It is a mirror. And what it reflects back at us is the uncomfortable image of a species that has been here longer, thought harder, and reached further than our textbooks are yet prepared to admit.

Who built it? How did they know what they knew? Where did that knowledge come from, and where did it go? What were they building next, before the silence fell?

I do not have those answers. I am not sure anyone does.

But I think we owe it to the mind that made that device — and to ourselves — to keep asking.