Anthropologists and Archeologists

Archaeologists and anthropologists have begun using AI tools to help with some tasks, especially scanning the land or images. For example, researchers now apply machine-learning and computer vision to aerial or satellite images to spot possible buried sites ^[1]. One study built an AI model that predicts where ancient tombs are likely to be, based on landscape features ^[1].

Drones and 3D scanners have also become common; the data they collect can be automatically processed. In one case an “AI-assisted workflow” (called AutArch) was able to take old publishing of artifact drawings and photos and automatically detect shapes like graves or pottery, turning years of drawings into data ^{[2] [2]}. This shows computers can speed up mapping and data recording.

By contrast, we found no examples of AI doing the more human-side tasks. Writing up discoveries or explaining cultural ideas still relies on people. Generative AIs (like ChatGPT) have been tried to help write code or summarize information ^[1], but not to replace a scholar’s voice.

And jobs like talking with communities about their health beliefs remain fully human – these tasks need trust and empathy. In summary, many technical parts of site surveying and data processing are increasingly supported by AI ^{[1] [2]}, but understanding and storytelling about people’s history still need human insight.

AI tools for these jobs are partly available today, but adoption depends on many factors. On one hand, there are free or cheap resources: for example, satellite images and GIS software can be used with ready-made AI libraries, and drones cost much less than they once did ^{[1] [2]}. In fact, researchers note that AI could “speed up, automate, and standardize” data collection, even freeing projects from tight budgets and time limits ^[2].

This suggests a clear benefit – if an AI tool helps find sites faster, people may want to use it. However, building and tuning these AI tools also takes money and skill. Many archaeology projects run on small grants, so leaders must decide if buying or developing AI is worth it.