My AI Agents Found an Unmapped Maya Site in Mexico

A traveler in Peru wondered how Hiram Bingham might have searched for ruins with modern tools, then turned to public lidar data covering parts of the Maya region. Lidar can map the ground beneath dense forest, revealing shapes hidden from ordinary satellite images. Forty-eight AI agents examined images made from roughly 1,200 square kilometers of data, following a guide to distinguish possible buildings and civic features from natural terrain and roads. They flagged numerous settlements; one hilltop site appeared especially promising, with a plaza, an oval pyramid estimated at 8-12 meters tall, a raised courtyard, platforms, and a hollow that could have served as a reservoir. Checks of the raw laser points supported the interpretation that these shapes were on the ground, not artifacts of vegetation.

The author, who is not an archaeologist, used the agents to research the location, prepare a Spanish-language report, and contact Mexico’s National Institute of Anthropology and History. INAH replied that its records contained no registered archaeological site there. The discovery is therefore not formally confirmed, and the site’s function and date remain speculative; the layout may indicate a small Maya civic center from the Classic period, roughly AD 250-900. Any field investigation requires INAH authorization, and the author offered to assist if officials pursue one. The account presents the rapid process as evidence that AI agents can help investigate questions beyond software and mathematics, arguing that curiosity, imagination, and access to computing may enable broader scientific exploration.