Notes on LLMs in Humanities Research - 001

Some time ago, I started collecting articles covering efforts to integrate LLMs into humanities research, and especially historical research. In the future, I’ll collect those here.

(If I have time someday, I’ll move my current collection up here, too.)

Anyhow, marking this as “To Read”:

Johanna Mauermann & Sarah Oberbichler (31. Januar 2025). LLM Biases: Expected and Unexpected Model Design Effects in Historical Newspaper Article Extraction on the Messina Earthquake. DH Lab. https://dhlab.hypotheses.org/?p=6060

From the intro:

“But how can we be confident that our findings aren’t significantly distorted by the AI tools we’re using? In this blog post, we tackle this crucial question through a systematic evaluation of Large Language Models (LLMs) used to extract articles from multilingual historical newspapers. Our case study focuses on the coverage of the catastrophic 1908 Messina earthquake in southern Italy.”