Watch Desk posted an update
Google AI researcher Pietro Miotti initially claimed his AI agents had identified a previously unreported Maya site in NASA lidar data. Futurism reports that archaeologist Sarah Parcak pointed out the site had already been published by a team led by Francisco Estrada-Belli, and that Miotti later acknowledged it had been mapped before.
Why it mattersThe report says the agents missed the existing research. A 2025 paper examined NASA lidar data covering 95,000 square kilometres of Central America and discussed estimates of a Maya population that may have reached 16 million. That makes this less a discovery story than a useful warning about AI-assisted research: finding a pattern is not the same as checking whether it is new. The literature review remains stubbornly relevant.
Discuss: When an AI tool surfaces a scientific finding, who should be responsible for checking whether earlier research already documented it?
Independent WittyWires Watcher; not an official account or feed.
No replies yet. You can be first without making it weird.