Project members
Jeremy Knox, Rebecca Eynon, Lulu Shi (Social Sciences Division).
Project summary
Developing and evaluate an AI system to support interdisciplinary discussions among postgraduate students by analysing their essays and generating prompts to foster productive dialogue and collaboration across diverse disciplines.
View final project report (PDF)
AI in Teaching and Learning at Oxford Knowledge Exchange Forum, 9 July 2025
Findings from projects supported by the AI Teaching and Learning Exploratory Fund in 2024–25 were presented at the AI in Teaching and Learning at Oxford Knowledge Exchange Forum at Saïd Business School on Wednesday, 9 July 2025.
Project team members each presented a lightning talk to all event participants, and hosted a series of small group discussions.
Follow the links below to view the lightning talk recording and presentation slides for this project.
View lightning talk recording (Panopto) - TO FOLLOW
View presentation slides (PDF)
Project case study
The AIID project developed a pilot AI-driven platform aimed at fostering interdisciplinary, in-person discussions among postgraduate students. The motivation behind this initiative stemmed from concerns about the diminishing quality of human interaction, as students are increasingly pushed toward solitary exchanges with automated AI systems. Instead of developing a typical AI interface where humans initiate interactions and the AI responds—as is common with Large Language Models (LLMs)—this project pursued the reverse: an AI designed to prompt human users.
The AIID project was developed to be used by a range of postgraduate students, preferably across different topic and disciplinary areas. Rather than being attached to the specific course of study, the AIID project aimed to provide a tool that would support additional forms of learning, where postgraduate students sought to discuss and develop their work in exchange with others, outside of formal class or tutorial time.
This project took a somewhat unconventional approach to using AI, which meant that existing, "off-the-shelf" AI tools were not always suitable. While a basic AI language model was used to generate discussion prompts, the processes of uploading writing samples and presenting the resulting outputs required a custom-built interface. Creating this bespoke system was the project’s central challenge, as such interfaces demand more time and effort to develop.
Future initiatives in this space should carefully consider the balance between custom-built and ready-made AI components. Weighing the trade-offs between speed of development and the need for tailored features is essential.
The project’s core idea—stimulating discussion—led to a number of related concepts, particularly around enhancing connections among students and staff across the University. Testers, especially students, also recognized the value of networking and discovering like-minded individuals. Although some features supporting this goal were explored, they were ultimately paused at this stage due to being underdeveloped and because they expanded the project's original scope.
Due to the timing of the development of the AIID platform, testing with students was limited. Many Masters students leave Oxford to conduct fieldwork for their dissertation at the end of Hilary term and are subsequently engaged in completing their writeup in Trinity. This should be considered early in the development cycle where testing is required. Nevertheless, the comments and feedback received indicated from the AIID project indicated a string interest in using AI for the purposes of connecting and engaging in discussion. Furthermore, student testers were forthcoming in offering interesting and relevant suggestion for future improvements.
