Skip to Content

A Lego Reconstruction of the World’s Earliest Computer

Before the birth of Christ the Greeks built a mechanical computer. Now an Apple engineer has made a functional Lego replica.
December 10, 2010

Here’s a brand new stop motion video of a reconstruction of the world’s first mechanical computer, directed by occasional Technology Review contributor John Pavlus. It’s entirely self-explanatory: watch it and read on.

One hundred years before the birth of Christ, when agriculture and the wheel was for most of human civilization the apex of technological achievement, the Greeks built a mechanical computer so sophisticated that it could add and subtract–all in the name of predicting the next lunar eclipse.

In 2010, Apple engineer Andrew Carol created, based on previous reconstructions of the so-called “Antikythera” mechanism, which was discovered in a shipwreck in 1901, a fully functioning Lego replica of the device. Like the original, it accurately predicts solar eclipses.

Its secrets are explained at length in a feature in Nature and its Wikipedia entry, which, not surprisingly for a device that is catnip for geeks, is as exhaustive as the plot exegeses of old episodes of Lost.

The Antikythera is such a marvelous device–Michael Edmunds of Cardiff University, who led the most recent study of the device, says it is more historically valuable than the Mona Lisa–that it continues to inspire great works of its own: first the historically faithful reconstructions of it, then the lego reconstruction, and now this stop motion video, which, like all stop motion, was an enormous effort in itself.

Here’s a sped-up, behind the scenes video of the shoot:

And if you want to really dig deep, Pavlus has conducted an interview with the creator of the Lego version of the Antikythera. It includes fascinating details about managing the friction generated by the more than 100 gears in the mechanism. Here’s Carol’s account of how it works:

It’s pretty simple; it’s all about ratios between the numbers of teeth on two gears meshed together. If one gear has 50 teeth and another has 25, that’s a 2-to-1 ratio – which means that turning the axle one full revolution on the first gear will multiply by two, because it turns the second gear twice as fast.

But the tradeoff is that when you make it go fast, you lose power. It’s fast, but it’s not strong, and vice versa – and those mechanical effects pile up quickly when you’ve got over 100 gears working together in exotic ratios. When I have to multiply by 127, it’s got to turn very fast, but with little power, which means that whatever amount of friction there is, I’ve effectively multiplied it by 127. So I had to put a lot of thought into designing the optimal layout of gears that would minimize the friction enough to make that kind of calculation physically work.

Finally, there’s Pavlus’s account of how he brought the video project itself together. For anyone interested in how to explain complicated technology to a lay audience, it’s quite a ride:

But how to actually execute that idea? Obviously some kind of animation would be necessary. Several people I consulted urged me to use computer graphics. But that felt wrong: Legos are wonderfully tactile, and I really wanted to highlight the machine’s intricate physical detail — to make you feel like you could literally reach out and touch the gears or turn the crank. CGI would feel too weightless and abstract — too perfect. Andy’s model was the quintessence of DIY hacking: he didn’t even diagram it out before starting to build it. I needed animation that was physical, craft-ey, and a little bit rough around the edges. Stop-motion was the clear choice.

For even more background on the Antikythera mechanism check out this illuminating video from 2008, produced for Nature by a director with decades of experience at the BBC.

Follow Mims on Twitter or contact him via email.

Keep Reading

Most Popular

Large language models can do jaw-dropping things. But nobody knows exactly why.

And that's a problem. Figuring it out is one of the biggest scientific puzzles of our time and a crucial step towards controlling more powerful future models.

OpenAI teases an amazing new generative video model called Sora

The firm is sharing Sora with a small group of safety testers but the rest of us will have to wait to learn more.

The problem with plug-in hybrids? Their drivers.

Plug-in hybrids are often sold as a transition to EVs, but new data from Europe shows we’re still underestimating the emissions they produce.

Google DeepMind’s new generative model makes Super Mario–like games from scratch

Genie learns how to control games by watching hours and hours of video. It could help train next-gen robots too.

Stay connected

Illustration by Rose Wong

Get the latest updates from
MIT Technology Review

Discover special offers, top stories, upcoming events, and more.

Thank you for submitting your email!

Explore more newsletters

It looks like something went wrong.

We’re having trouble saving your preferences. Try refreshing this page and updating them one more time. If you continue to get this message, reach out to us at with a list of newsletters you’d like to receive.