Schiavitù e invented archives. Intelligenza Artificiale generativa nella costruzione del Database on the Slave Trade

Abstract

This essay examines the construction of the Database on the Slave Trade between the Mediterranean and the Atlantic (15th–16th centuries) as a case study to explore the transformative role of generative Artificial Intelligence tools in historical and archival research. Positioned within the framework of histoire sérielle, the project addresses methodological challenges related to the datafication of sources on Mediterranean and Atlantic slavery, showing how invented archives function as epistemic spaces where memory, computation, and interpretation intersect. The use of LLMs (Large Language Models) for data extraction, encoding, and modeling raises critical issues concerning scientific validity, algorithmic opacity, and the indispensable role of human oversight. The essay offers a critical reflection on the integration of Digitality and Historiography, highlighting both the limitations and potentials of emerging technologies in transforming historical documents into queryable cybertexts and in fostering the collaborative construction of historical memory.

La ricerca si e avvalsa di un finanziamento dell’Unione europea – Next Generation EU – missione 4, componente 2, investimento 1.1, nell’ambito del programma PRIN-PNRR. Il titolo del progetto e: A Database on the Slave Trade between the Mediterranean and the Atlantic (15th-16th centuries).