Did AI Really Find Giant Pillars Under the Pyramids?

Tthey did not point a normal radar beam through 600 metres of limestone and photograph pillars.

Their method derives from Biondi and Malanga’s 2022 paper on what they call SAR Doppler tomography. Conventional SAR primarily images surfaces, something their own paper explicitly acknowledges.

Their technique instead attempts to infer internal structures by measuring extremely small surface motions, supposedly caused by seismic vibrations, wind and other sources, and reconstructing what produced those vibrational patterns.

That distinction is fascinating because some criticism has been slightly too easy.

Radar specialist Lawrence Conyers correctly noted that ordinary radar cannot penetrate hundreds of metres of rock, but the researchers are claiming something more indirect than conventional ground-penetrating radar. Their method is closer in concept to trying to infer what is inside an object from how the object vibrates.

The real question therefore becomes whether their reconstruction method can reliably turn those surface motions into a detailed three-dimensional map hundreds of metres beneath Giza. That spectacular capability has not been independently demonstrated.

And AI adds another layer of confusion. AI-generated or heavily rendered images have accompanied the story online and attracted criticism, but the claimed discovery itself was presented as coming from SAR data and specialised signal processing, not an AI system autonomously discovering Egyptian architecture.

Some very convincing pictures circulating online are therefore visualisations of an interpretation, not photographs of enormous stone columns patiently waiting underground.

There is also a superb comparison available: technology really has discovered a major hidden structure inside a Giza pyramid.

In 2017, the ScanPyramids project detected a previously unknown void at least 30 metres long inside Khufu’s Great Pyramid using cosmic-ray muons. It was independently detected with three different muon systems and published in Nature. That gives us a useful example of what extraordinary non-invasive archaeological evidence looks like when several independent methods converge.

Even better, another international team is currently developing muon tomography specifically for Khafre’s pyramid, precisely to look for unknown internal voids. That project offers an obvious future test of at least some claims concerning the pyramid itself, although muography from the planned detector positions is not designed to magically settle kilometre-deep underground-city claims.

That gives us a killer final direction:

If those eight enormous cylinders are merely artefacts created by an experimental reconstruction method, independent scanning should eventually make them disappear.

But if another technique detects something in the same places, what exactly is underneath Khafre?