NTU College of Science E-Newsletter, Issue 59
2024 Symposium on AI and Humanities Research

The rapid development of AI is reshaping the future of humanities research and education. On November 29, 2024, Professor Su-Ling Yeh and her AI research team hosted the 2024 International Symposium on AI and Humanities Research, which aimed to explore the multifaceted relationship between AI and the humanities. This event focused on how the humanities can both enhance and critique AI development, the role of AI in humanities research, and its potential to enrich and transform our understanding of history, literature, art, and philosophy. The symposium welcomed eight renowned international scholars and four leading Taiwanese experts, attracting over 150 participants from both local and global academic communities. Video recordings of the sessions are available on the symposium’s official website.

​

During her opening remarks, Professor Yeh expressed gratitude for her enriching experience at the National Humanities Center in the U.S. the previous year, which inspired her to invite several fellows to Taiwan for this event. The invited speakers’ willingness to participate during the Thanksgiving holiday highlighted their enthusiasm for advancing the integration of humanities and AI. The symposium created a vibrant atmosphere of academic exchange, likened to a digital humanities ballroom dance where participants explored various rhythms and styles of interaction between AI and the humanities. Whether analyzing historical data with a two-step, waltzing with literary works, or breakdancing to philosophical concepts, everyone found their unique rhythm.

​

Professor Yeh introduced the symposium by presenting the TAIHUICAS Project (Taiwan Humanities Conversational AI Knowledge Discovery System), which leverages language models to assist humanities scholars in Taiwan-focused research. This system enables users to interact with data from linguistic and cultural archives related to Taiwan. Professor Tsung-Han Tsai expanded on the technical aspects of TAIDE, a locally developed Taiwanese language model, highlighting the computational resources needed to operate it efficiently on personal computers.

​

Professor Rachel Levy emphasized the importance of collaboration among humanities scholars, artists, and scientists to explore new approaches to data science and AI-driven innovation. Professor Shaoling Ma discussed narratives from Cold War-era Taiwan that reveal complex interactions between economics and culture, providing insights into how AI can better reflect cultural contexts. Professor Devin Fergus examined the potential for AI to democratize lending, focusing on challenges such as regulatory constraints, bias, fairness, and the opacity of large language models. Professor Ron-Chiang Chen highlighted AI’s capacity to uncover individual stories within historical records, offering methods for addressing gaps in archives and artifacts.

​

Professor Po-I Hung discussed the challenges and opportunities of using AI in field research, including issues such as informed consent, the loss of firsthand experience, and contextual information. He highlighted how AI can reduce data distortion, preserve research integrity, and reaffirm the critical role of humans in knowledge production.

​

Professor Hung-Yi Lee explored the possibility of AI providing real-time feedback and grading student assignments, advocating for transparency in grading systems while anticipating potential student strategies to bypass such systems. Professor Sally E. Hadden shared insights on using AI in classroom settings, cautioning against unawareness of missing information in datasets due to unrecorded or unprioritized data. Professor Da-Lun Huang discussed how underrepresented data are marginalized in AI systems due to algorithmic and training constraints.

​

Professor Lynn Mie Itagaki addressed power dynamics related to authenticity claims, re-evaluating the concept’s significance in the AI era and encouraging participants to explore the motivations behind various actions. Professor Katherine Davies offered philosophical insights into thought and cognition, urging the audience to reflect on how AI intersects—or fails to intersect—with these concepts.

​

Cross-disciplinary themes such as accessibility, authenticity, bias, fairness, and missingness emerged throughout the symposium. Attendees appreciated the breadth of topics, which allowed them to connect their humanities work with AI applications. The speakers collectively underscored the importance of the humanities in shaping AI development and highlighted the potential of AI to support humanities research through improved accessibility and accuracy. This event showcased diverse AI applications in the humanities, emphasizing the value of interdisciplinary collaboration in advancing these technologies and inspiring future academic exploration.

​

Acknowledgment was given to the NTU College of Science for providing the venue and to Professor Yeh’s lab for funding and manpower, which were instrumental in the success of this conference.