Currently reading Magnani & Clindaniel (2025) on “Artificial Intelligence and the Interpretation of the Past”

No matter how sick I am of GenAI, this is a cool paper! Magnani & Clindaniel (2025) look at lots and lots of GenAI generated texts and images — both generated with a generic prompt and with one specifically asking for scientifically correct representations — depicting Neanderthal behavior and compare it to what science actually says.

Apparently, the ideas we have about how Neanderthals lived have changed substantially over the last 100 or so years, so one can compare those GenAI generated representations with different periods of knowledge about them. And they find that GenAI reproduces “dated knowledge and cultural anachronisms“, and also gender bias in that almost all Neanderthals are depicted as muscular men, with no women appearing and hardly any children. Overall, images are much more stereotypical than texts (that are still stereotypical, though).

But what I find especially interesting is the discussion of the influence of available training material for GenAI models, and how that might influence future understanding of the past:

“Our current research suggests that the way we structure and make information available will directly influence AI output and, by extension, the way we imagine the past. Moving forward, data policies will inform the way archaeological material is written about and visualized. Open access structures the types of information that are accessible for scholars to conduct computational analyses and is equally likely to affect which information is used to train artificial intelligence programs. Institutions with robust funding to allow articles to be widely read will continue to have a greater hand in determining how things are represented. If articles from certain eras or from specific subfields are more likely behind paywalls, they will be less likely to feed into materials generated using AI. Academic publishing practices may ultimately work to undermine public knowledge.”

And of course, the impact of the findings in this article is much larger than just how we think about Neanderthals (that I don’t tend to think about a lot), but ultimately applies to all GenAI-generated information.

Very cool article that I will definitely bring into discussions with colleagues!

Magnani, M., & Clindaniel, J. (2025). Artificial Intelligence and the Interpretation of the Past. Advances in Archaeological Practice, 1-16.

Quick dive spot check in Limhamn on Saturday! Divespot still there!