Digital humanities și revoluția inteligenței artificiale

Digital Humanities (DH) is a scientific domain, albeit an elusive one. Broadly, DH studies digital cultural objects in a computational way. There is no commonly accepted definition. The challenge in defining DH stems from the wide variety of disciplines involved, like cultural heritage, linguistics, literature, digital archaeology, history, arts, philosophy, etc. that all have their own methodology and tools. A common misunderstanding is that the DH is concerned only with the conversion of physical objects like manuscripts, maps, cultural artefacts, sounds, into a machine-readable format. But this is only the first step: ob¬taining the object of study. The most challenging part is the computational processing of the digital objects by operations such as interrogation, editing, annotating, visualization, Data Mining, automatic classification, clustering, pattern recognition, information extraction, etc. Only a decade ago, all of these could have been performed exclusively by scholars with computer science background, but now the tendency is to develop user-friendly tools, or use prompt engineering for large language models, to make it easy for humanists to benefit from these computational analysis methods. In the last half of the decade, we witnessed a fourth industrial revolution, with the fulminating rise of Artificial Intelligence technologies. In this dynamic context, the role of DH scholars increases. Smart, creative, adaptive, and visionary digital humanists are needed to use AI systems to benefit people. In short, we need digital literacy on a large scale and adapting to continuous learning. An example of such good practice is the array of dissertation topics proposed by students graduating the Digital Humanities master Program at FLLS, like the computational analysis of the discourses of the last four Romanian presidents, the automatic detection and classification of mental illnesses from social media posts, the diachronic analysis of semantic shift of gender representation in corpora, the comparative semi-automatic analysis of literary cur¬rents, the automatic classification and sentiment analysis on a textual corpus of dreams, the comparison of human and AI-generated fanfiction texts, or testing the verbal creativity of large language models and compare it with human performance, etc. Pannapacker predicted as soon as 2012 that “It won’t be long until the Digital Humanities are, quite simply, The Humanities”. That moment has already passed.