Scientists are increasingly frustrated by amateurs using AI to claim "discoveries" in professional fields.
Google researcher Pietro Miotti recently claimed he used a team of AI agents to locate a potentially unmapped Maya site in Mexico. However, this claim was quickly debunked as the site had actually been documented in an academic journal a year prior. This incident has sparked significant criticism from the archeological community, highlighting the growing tension between non-specialists leveraging powerful AI tools and experts who possess the necessary domain knowledge to interpret findings correctly.
Miotti conceived the idea while traveling in Peru, considering how historic explorers would utilize modern technology. He utilized lidar, a laser-based mapping technique capable of creating detailed 3D images, to analyze NASA data for archaeological sites. Since Peru lacked comprehensive public lidar coverage, he shifted his focus to the Maya region. His team deployed 48 AI agents to scan the data for undocumented locations.
The agents identified several promising areas, including a "major village" that appeared unregistered in existing databases. Miotti, who conducted the research as a personal side project without utilizing Google resources, contacted Mexico’s National Institute of Anthropology and History (INAH) to verify the findings. Using AI to translate his correspondence into Spanish, he was informed by INAH that there was no record of a site at that specific 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 in a blog post.
Miotti clarified that he is not an archeologist and emphasized that the finding requires verification through physical fieldwork. Despite this, he hailed the project as evidence of a scientific revolution and advocated for the entire planet to be mapped with lidar so AI could assist in identifying remaining sites. Had he possessed regional expertise, however, he likely would have identified the issue earlier, as the site had been previously published in an academic journal.
"I didn't expect it to get this much attention," Miotti told The Verge.
Archeologists quickly pointed out the oversight on social media. Sarah Parcak, a professor at the University of Alabama at Birmingham, stressed the importance of cross-referencing findings with existing literature. "This is why we have peer review," Parcak wrote, noting that a lack of registry in government databases does not equate to the site being unknown or unmapped. Other experts echoed this sentiment, criticizing the project for spreading misinformation and detracting from the decades of work archeologists have invested in digital tools.
Miotti explained that he was unaware that a site not being registered was not synonymous with being unmapped. He admitted that his AI tools missed the existing documentation because archaeologists rarely publish exact coordinates, making it difficult for agents to search for figures in papers rather than database entries. He promised to update his post to credit the experts who pointed out the error.
Miotti suggested that AI agents could still serve a valuable role by helping experts manage large datasets. He noted that several archaeologists and PhD students have reached out to him, expressing excitement about the workflow and its potential application to their own survey data.
"Our 'problem' is not discovering new sites," Elizabeth Graham, a professor of Mesoamerican archaeology, told The Verge. She noted that local communities are often already aware of sites in their areas and that the lack of funding for excavation and mapping remains the primary obstacle to further progress.
"Just spotting something does not add much to our knowledge."
This controversy mirrors broader issues within the scientific community regarding the use of AI. While tools like those developed by OpenAI are becoming increasingly capable, they often fail to engage with existing research, potentially leading to claimed breakthroughs that have already been achieved or a general failure to credit prior work. The amateur use of these tools shifts the burden of verification and contextualization onto actual experts.
Ivan Šprajc, a professor of archaeology at the Slovenian Academy of Sciences and a coauthor of the paper documenting the site, expressed disappointment with the agents' performance. He argued that archeologists have used advanced remote sensing for years and that spotting a feature via lidar does not constitute a significant contribution. He also warned that the rising number of AI "discoverers" could overwhelm professionals with unverified claims and potentially attract looters to sensitive sites.
The Editorial Staff at AIChief is a team of professional content writers with extensive experience in AI and marketing. Founded in 2025, AIChief has quickly grown into the largest free AI resource hub in the industry.