The warning of a historian, related to artificial intelligence: we must not try and overcome the “machine”, but be as different as possible

In this era of artificial intelligence historians still have their place, a researcher who, in an article published in a specialized magazine, warns that, if we leave historical reports to AI, there is a risk that people's sufferings will be omitted from many testimonies.

Jan Burzlaff is an expert in the history of Nazi Germany at Cornell University and-when he asked Chatgpt to summarize the experiences of the Holocaust survivors-found that the chatbot had omitted intimate and essential details.

He wrote in the magazine Rethinking History a research with the title “Fragments, not prompts: five principles for writing in the age of ai”, all starting from AI asking Chatgpt to summarize the testimonies of Holocaust survivors.

“In the testimony of Luise D., a Holocaust survivor, the AI ​​failed to give her mother, who cut her finger to give the child, who was dying, a few drops of blood, to keep her alive,” says the historian, quoted by Phys.org.

This omission demonstrates why historians remain indispensable in the era of artificial intelligence, he says.

He says that there is the crucial need to capture the emotional and moral complexity behind world events, and this can only be done by historians “in flesh and bones”.

“As tools such as Chatgpt are increasingly entering education, research and public discourse, historians have to face what these systems can do, but also with what they cannot do. These chats summarize, but do not listen; reproduce, but interpret; excel in coherence, but fail in the face of the contradiction. The machine “, but to do anything so that we do not look at it at all.”