Using generative AI as part of historical research: three case studies

Using Generative AI in Historical Research: Three Case Studies

The integration of generative AI in historical research has opened new avenues for discovery and analysis. Despite challenges in educational settings due to AI misuse, these tools demonstrate potential in scholarly fields through innovative applications. Here, three case studies show how AI is reshaping historical research.

Case Study #1: Transcribing and Translating Early Modern Italian

GPT-4o’s application in transcribing early modern Italian text from Urbano Monte’s world map demonstrates its proficiency. The AI accurately transcribed and translated Monte’s text, providing a useful resource for historical research. Although a few minor errors were present, its output was largely reliable. This task, traditionally requiring extensive training, was completed efficiently, showcasing the potential of AI in handling historical texts.

Case Study #2: Analyzing an 18th Century Medical Text

OpenAI’s o1 model transformed the understanding of a unique 18th-century Mexican medical manuscript. The model successfully transcribed and translated the text, providing a detailed historical and iconographic analysis. This analysis offered insights into the manuscript’s context, emphasizing motifs such as sun-faces and serpentine lettering, which align with early modern Iberian and colonial iconography.

Case Study #3: Generating New Historical Interpretations

The comparison between OpenAI’s o1 model and Anthropic’s Sonnet 3.5 illustrated AI’s capability in generating historical interpretations. The AI provided novel frameworks for analyzing figures like Francis Galton and William James, suggesting innovative historical arguments. These insights can supplement a historian’s perspective, augmenting their research capabilities.

Conclusion

The applications of generative AI in historical research reveal its dual role as both a facilitator of new insights and an aid in traditional tasks such as transcription and translation. While concerns about AI replacing human creativity exist, the current trajectory suggests that AI will serve as a complementary tool, enhancing historians’ ability to interpret the past. As AI technology advances, its role in historical research and analysis will likely become even more significant.

For further insights and in-depth analyses, consider subscribing for additional content and special features.