Armed with a team of AI agents, Google researcher Pietro Miotti claimed earlier this week to have discovered a potentially unmapped Maya site in Mexico. There was just one problem: The site had already been discovered and was published in an academic journal last year. The incident, which sparked criticism among archeologists online, underscores a growing tension in science as amateurs wielding increasingly powerful AI tools parachute into specialist fields, often without the expertise needed to understand what they are looking at.
Local communities are often already well aware of sites in their area.
Miotti said the idea to use AI to search for archeological sites came to him while traveling in Peru, when he began wondering what historic explorers might do with access to modern technology. He settled on lidar, a laser-based mapping technology that can produce detailed three-dimensional images of the environment, and set a team of 48 AI agents to comb through NASA lidar data for signs of undocumented sites. As “Peru isn’t well covered” by public lidar data, Miotti wrote, he shifted his focus to the Maya region.
The agents flagged numerous promising locations, including “one major village [that] doesn’t seem to have been registered” in existing records. Miotti, who described the research to The Verge as a personal side project done without any Google compute or resources, said he then reached out to Mexico’s National Institute of Anthropology and History (INAH) to check whether the apparent discovery was already known, including through unpublished research. Using the agents to translate his correspondence into Spanish, Miotti said INAH told him it had no record of a registered site at that location.
“It took less than 72 hours to go from a random thought while traveling to an official reply from the Mexican government. This is by far the best time ever to be alive,” Miotti wrote on his blog.
Miotti stressed that he is not an archeologist and emphasized that the finding would need to be verified with fieldwork. Nevertheless, he hailed it as proof that “we’re on the brink of a revolution across every science” and called for the “entire planet” to be mapped with lidar so AI agents could “help us find the rest.” Had he been an archaeologist familiar with the region, however, Miotti might have spotted a problem: The site was already known, and had been documented in an academic journal more than a year earlier.
“I didn’t expect it to get this much attention.”
Archeologists were quick to point out the oversight on social media. In a series of posts on Bluesky, Sarah Parcak, an archeology professor at the University of Alabama at Birmingham, stressed the importance of carefully checking findings against existing research. “This is why we have peer review,” Parcak wrote, adding that just because the Mexican government doesn’t have something in its database doesn’t mean it is unknown or unmapped. Other archeologists echoed her criticism, slamming the project for spreading misinformation and taking credit away from archeologists, who have been using advanced digital tools in their work for decades.
“This was a personal project that I was excited to share,” Miotti told The Verge. “I didn’t expect it to get this much attention.” He added that as a non-archaeologist, he had not realized a site not being registered wasn’t the same as it not having been mapped, adding that expert knowledge “is needed to verify what the agents find.” He also addressed the limits of his AI tools: “My agents and I missed it. Archaeologists rightly don’t publish the coordinates of sites, so checking whether a site is already known means comparing images against figures in papers, not searching a database. That’s much harder for agents, and it’s something these tools need to get better at. When people pointed it out, I updated the post and the blog as soon as I could to credit their work.”
Nevertheless, Miotti said this was an example of how agents could be used to help experts sort through large amounts of data. He said he has been encouraged by the archaeologists and PhD students that he said were reaching out to him “excited about the workflow and wanting to know how it works and whether it could help with their own survey data.”
“Our ‘problem’ is not discovering new sites,” Elizabeth Graham, a professor of Mesoamerican archaeology at University College London, told The Verge. “Most of us working in the Maya area know about tons of sites not explored or excavated or mapped, most of these found on foot,” she said. Graham also critiqued the use of the word “discover,” noting that local communities are often already well aware of sites in their area.
While she credited Miotti with doing a decent job of using AI to map the mounds from lidar data — “the interpretation of lidar data can be tricky” — Graham argued the bigger obstacle is funding for research. “The fact is, there just isn’t the funding to excavate or map sites as there was in the past, so in a way it doesn’t matter how many more are ‘discovered.’”
“Just spotting something does not add much to our knowledge.”
The controversy reflects a broader problem with AI in science, perhaps most visible with OpenAI’s recent advances in mathematics. While AI tools are increasingly capable of producing genuinely impressive results, they don’t always engage with the existing research those findings build upon. The result can be claimed breakthroughs that have, as is the case here, already been made by someone else, or a general failure to credit earlier work. Even when the findings are actually new, they can lack context a specialist would bring. Ultimately, amateur use of AI tools often ends up shifting the burden of establishing what has actually been found and why it matters to actual experts, work that would ordinarily be part of making a scientific discovery in the first place.
Ivan Šprajc, a professor of archaeology at the Research Centre of the Slovenian Academy of Sciences and Arts in Ljubljana and a coauthor of the paper documenting the site, told The Verge he wasn’t particularly impressed by the agents’ work. Archeologists have been using lidar and “other sophisticated remote sensing tools for quite some time,” he said, adding that “just spotting something does not add much to our knowledge.” He told The Verge that while it is commendable that Miotti checked in with INAH, doing so is “not sufficient” as the institution does not have access to all possible information. “The fact that INAH has no official knowledge of a particular cluster of archaeological features does not necessarily mean that it has not been spotted or even explored by archaeologists.”
Šprajc said that he worries the growing use of AI tools by amateurs could overwhelm archeologists with claims of new discoveries. “Just today I have received an email from another AI ‘discoverer,’ so I am a little bit concerned about being flooded by this kind of message,” he said. He also raised another potential risk of amateurs publicizing the locations of archeological sites: attracting looters. “But as long as these findings do not lead to looting, I don’t see a major problem,” he added.
