Artificial Intelligence (AI) is bringing significant changes to various fields of knowledge, and the Digital Humanities (DH) are also experiencing this shift. Advanced computational techniques are transforming traditional research methods, allowing for the analysis of large datasets, uncovering patterns that were previously unattainable and answering new research questions. However, the attempt to create human-like intelligence in machines often views intelligence as a static target, ignoring its fluid and evolving characteristics. This perspective also applies to the extensive datasets used to train these models, failing to consider the nuanced and context-specific aspects of language that are essential to humanistic studies.
On 17 June, 2024, the University of Reading Digital Humanities Community of Practice hosted its inaugural conference to explore these issues and discuss the benefits, drawbacks and challenges of the interaction between DH and AI. The conference focused on three of Reading’s research strengths: Cultural Heritage, Synthetic Media and Ethics; and Teaching.
Dr Mara Oliva, Associate Professor in US History and Digital Humanities Champion, welcoming delegates. Photo credit: Professor James Ferryman.
Dr Barbara McGillivray (King’s College London) delivered a thought-provoking address entitled: “Coding Culture: A Journey at the Nexus of Artificial Intelligence and Digital Humanities”. Her insights into the symbiotic relationship between coding and culture emphasised how AI can be harnessed to decode and encode cultural narratives, thus enriching our understanding of diverse cultural phenomena. She highlighted case studies demonstrating AI’s potential in uncovering hidden patterns in historical textual collections. Her address underscored the importance of ethical considerations and the need for interdisciplinary collaboration to navigate the complexities at this nexus.
Dr Barbara McGillvray taking questions after her keynote address. Photo credit: Professor James Ferryman.
The sessions on cultural heritage showcased how AI tools are revolutionising the preservation and interpretation of historical artifacts and intangible cultural heritage. Discussions included the use of machine learning to restore, digitise and catalogue manuscripts, the creation of virtual reconstructions of archaeological sites, and the deployment of AI in curating museum collections. While these innovations can preserve cultural heritage and make it more accessible to the public, fostering a deeper appreciation and understanding of our shared history, concerns were raised about the uncritical use of datasets.
The exploration of synthetic media in the conference highlighted both its creative potential and its ethical dilemmas. AI-generated content from deepfake videos through tracing the history of digital photo colourisation, to AI-composed music, offers new artistic avenues but also poses significant risks related to misinformation and intellectual property rights. The sessions encouraged a balanced perspective, advocating for responsible use in media creation while stressing the need for robust regulatory frameworks to mitigate potential harms.
Dr Jon Weinbren (University of Surrey) on Synthesising the Obraz/ Stanislavsky, Active Analysis and AI-Assisted Image Creation. Photo credit: Professor James Ferryman.
The ethical considerations surrounding AI in Digital Humanities were a recurrent theme throughout the conference. Discussions ranged from data privacy issues to the biases inherent in AI algorithms. Speakers emphasised the necessity of developing ethical guidelines and best practices to ensure that AI applications in the humanities are fair, transparent, and accountable. The sessions underscored the role of human oversight in maintaining ethical standards and the importance of diversity in AI development to prevent cultural biases.
The inclusion of AI in humanities education was another focal point of the conference. Presenters shared innovative pedagogical approaches, such as using AI-driven tools to enhance language learning, facilitating interactive historical simulations, and employing data analytics to personalise learning experiences. These sessions highlighted the potential of AI to transform teaching methodologies, making education more engaging and tailored to individual needs. However, they also called attention to the challenges of digital literacy and the need for educators to be adequately trained in these new technologies.
The conference served as a crucial forum for advancing interdisciplinary collaboration and fostering a thoughtful approach to harnessing AI’s potential in the humanities. Challenges such as: ethical considerations, technological limitations and the need for interdisciplinary collaboration were identified as crucial areas for further investigation.
Mara Oliva is Associate Professor in Modern US History and the University’s Digital Humanities Community of Practice Lead and & Digital Humanities Champion. For more information please visit the Digital Humanities Hub.