Cambridge AI in Teaching Forum
The first event of the Cambridge AI in Teaching Forum - a gathering of those actively involved in teaching at Cambridge and interested in using AI or the impact of it on teaching and teaching practice.
The Forum was opened by Dean of Cambridge Judge Business School, Professor Gishan Dissanaike, followed by David Stillwell, Professor of Computational Social Science at Cambridge Judge Business School who delivered a session on current and future AI capabilities, and the three stages of AI in teaching: Denial, Avoidance, and finally Acceptance.
We welcomed academics from across the University to deliver five rapid-fire presentations on the use of AI in teaching and concluded the Forum with an audience discussion facilitated by Professor Stillwell.
Below you can find more details on each of the presentation as well as access to the slides used.
Cambridge AI in Teaching Forum - September 2025
Keynote Presentation: Current and Future AI Capabilities, and the Three Stages of AI in Teaching: Denial, Avoidance, and finally Acceptance
David Stillwell, Professor of Computational Social Science at CJBS
d.stillwell@jbs.cam.ac.uk
The session explored how generative AI is rapidly reshaping work, creativity, and education. It unpacks its current capabilities and future directions, reflects on the opportunities and risks it poses, and considers how educators can rethink assessment and prepare students for human–AI collaboration
Rapid-Fire Presentation 1: A Very Short Introduction to Using Generative AI (GenAI) for Teaching and Researching in the Sciences and Humanities
Dr Anthony Harris, Fellow and Tutor (Clare Hall). Director of Studies for Computer Science (Emmanuel College), Computer Science and Technology and ASNaC (Faculty of English)
awh28@cam.ac.uk
A few practical examples/case studies of the use of GenAI for teaching and researching in the humanities and computer science, based on my experience(s) of employing it over the past few years.
Rapid-Fire Presentation 2: Google Colab Meets AI and What Next?
Jon Bonsor-Matthews, Teaching Associate, Computing, Department of Engineering
jpm66@cam.ac.uk
Teaching Python. Issues with Google Colab AI providing code solutions before allowing students to try themselves. Moving to Jupyter Lab – challenges with this. V. brief intro to our ongoing thoughts for embracing teaching of AI Code Gen later in our course.
Rapid-Fire Presentation 3: Lessons Learned (and Uncertainties Remaining) from Incorporating Generative AI in a New Undergrad Data Analytics Course in Geography
Dr Becky Dell, Assistant Professor of Glaciology in the Department of Geography and Scott Polar Research Institute
rld46@cam.ac.uk
Dr Matt Osman, Assistant Professor of Climate Science in the Department of Geography
mo549@cam.ac.uk
We share our experiences running a new 3rd-year course in Geography, where students (without assumed maths/coding backgrounds) were encouraged to view AI as a troubleshooting resource whilst developing code for environmental data analysis. We reflect on how student reliance evolved throughout the year, and what we learned about balancing legitimate applied skills and self-sufficiency in data analytics with rapidly shifting ‘real-world’ AI norms and capabilities.
Rapid-Fire Presentation 4: Quiz Question Generation Using Co-pilot Studio: Embedding Best Practices and Criticality into Design
Stewart Utley, Learning Designer for Cambridge Online Education (CUPA)
stewart.utley@cambridge.org
We’ll be briefly introducing how we utilised Co-pilot Studio to generate a range of ‘Question bots’ capable of supporting Learning Designers and Subject Matter Experts in generating quiz questions for course content. We’ll lift the hood to explore some of the logic and prompting processes that make them work, consider how we integrated assessment best practices, and a very brief insight into a small study on their perceived quality’.
Rapid-Fire Presentation 5: Using ChatGPT to Visualise Lecture Concepts
Dr Daria Frank, CATAM Assistant Director, Fellow, DoS in Mathematics and Tutor at Corpus Christi College
dt346@cam.ac.uk
In this session, I will present my practical experience using ChatGPT to create Python-based simulation tools that help students better understand lecture concepts. I will focus on examples from Fluid Dynamics lectures, though similar applications are possible in other teaching areas.