The granite used in the Great Pyramid was transported from quarries located more than five hundred miles to the south. The ceiling of the King’s Chamber, for example, consists entirely of massive granite blocks, and its surfaces show details that have been studied closely for years.
In the quarries at Aswan, long, curved, and evenly spaced grooves have been cut into one of the hardest rocks on Earth. The cuts are so smooth and regular that many people who see them for the first time assume a modern machine must have made them. The markings are thousands of years old, however, and they have become the subject of a long-running debate about how they were produced.
When the rock surfaces were examined with AI analysis, the results contradicted the standard explanation found in history books. If the analysis is correct, the accepted story about how these grooves were made may be incomplete.
Granite ranks high on the hardness scale, and the dark dolerite veins running through it are even harder. Modern construction crews still cut this stone using diamond-tipped saws and heavy industrial equipment. The material does not yield easily to hand tools, and steel blades quickly become damaged when they come into contact with it.
The grooves in the Aswan quarry are clean and uniform rather than chipped or hammered. Their walls are consistent, and the spacing between them is regular to a degree that is difficult to explain by hand. Hundreds of these markings appear across quarry walls, floors, and abandoned blocks, and they repeat the same unusual smoothness.
The official explanation has been that workers used dolerite balls, which are harder than granite, to pound the rock into shape over long periods. According to this account, thousands of laborers rotated through shifts, slowly wearing away the stone one grain at a time until the desired forms emerged. This version of events has been taught in schools, printed in textbooks, and repeated by guides in the quarry for decades.
Researchers eventually tested this theory directly. They collected dolerite balls and asked experienced workers to pound granite blocks the same way the books described. The results did not match the ancient markings. The hammering produced rough, uneven surfaces. Some strikes cut deeper than others, and the edges broke and spalled. No smooth curves or regular spacing appeared. The surfaces looked damaged rather than shaped. A more recent study measured the rate of stone removal using this method and concluded that hammering alone cannot account for the actual work.
Since the traditional theory failed the test, researchers tried a different approach. They removed human assumptions from the process and gave the problem to an AI system called Grok, developed by Elon Musk’s company. The research team deliberately withheld historical context. They did not mention Egypt, pharaohs, quarries, or pyramids. They supplied only raw three-dimensional scans of the grooves, along with a large library of tool marks from both ancient and modern equipment.
The AI measured how much the grooves varied from one another. Human hands naturally produce inconsistency over time. Fatigue, changes in grip, and shifting angles create measurable variability that experts expect to fall between 12 percent and 20 percent. Across hundreds of grooves in different parts of the quarry, the AI found variability of only about 4 percent. That level of consistency is difficult for any human hand to maintain over many repetitions, while a machine can sustain it indefinitely.
The AI also examined the angle at which the cutting tool entered the stone. Across hundreds of marks, the entry angle remained within a few degrees of uniform. It then analyzed narrow vertical shafts in the quarry, some extending thirty to fifty feet deep and only a few feet wide. The same clean grooves appear on the walls of these shafts. There is not enough room inside them for a worker to swing a heavy dolerite ball, yet the markings are identical to those in the open quarry. The AI concluded that a tool requiring a backswing could not operate in such a space, while a tool that cuts straight down could work in both open and confined areas.
Microscopic examination revealed further details. Hammered granite typically shows a shattered surface with micro-cracks radiating into the stone. The Aswan cuts instead show clean, sharp boundaries with little damage beneath the surface. The force appears to have been concentrated along a precise line rather than spread across the rock. Fine spiral patterns run along the cuts, consistent with a rotating tool moving in one direction, and they all rotate the same way. Small traces of heat were found at contact points, and tiny hard grains similar to those used in modern industrial cutting discs were present as well.
The AI then arranged the grooves by age. In most civilizations, craftsmanship improves over time, with later generations producing more refined work. In Aswan, the pattern runs in the opposite direction. The oldest cuts are the most precise, and later work is rougher and more variable. The trend appeared across the full data set and was statistically significant.
There is a site in Aswan that highlights this question directly. The unfinished obelisk is the largest monument of its kind ever attempted in Egypt, more than 130 feet long and weighing close to one thousand tons if stood upright. It remains where workers left it, still attached along one side to the bedrock. The official explanation is that a crack appeared in the granite and the project was abandoned. Cutting marks continue beyond the crack, however, suggesting work went on after the defect was visible.
The same kinds of grooves appear at Giza. In the valley temple near the Sphinx, granite blocks weighing hundreds of tons are fitted together so precisely that a knife blade or a sheet of paper cannot pass between them. Under magnification, these surfaces show the same smooth separation and faint rotation marks seen in the Aswan quarry. The AI compared the markings from the two sites, separated by hundreds of miles, and found that they differed by less than 5 percent, as if the same tool and technique had been used in both locations.
When asked to identify the closest match among all known tool marks, the AI settled on a rotating industrial cutting process with a confidence level near 94 percent. That category of machine, according to the conventional timeline, was not developed until the twentieth century. No machine has ever been discovered in Egypt, however, and no diagrams or power sources have been found. Traditional scholars continue to argue that human labor, sharpened copper, hard stone, and sufficient time could explain the work. The hammering experiments produced marks that do not resemble the ancient cuts, and the oldest work remains the most precise. The AI analysis, based only on geometry and tool-mark data, points toward equipment that should not have existed at that time. The evidence does not settle the question, but it raises doubts about the standard account.
