Welcome back to the AI Commons. This fortnightly newsletter is funded by the Institute for Teaching and Learning and run by Mark Carrigan and Eva Parr. We're exploring how colleagues are integrating AI into teaching and research, with the aim of supporting learning across the university so we can approach this change in thoughtful, creative and collegial ways. If you'd be interested in contributing or have suggestions for future topics please contact [email protected]

📝 In this issue, Peter Kahn (MIE) argues that addressing AI in assessment isn't something academics can manage alone and that shared infrastructure matters more than individual fixes:

It would be possible for the University of Manchester to require, say, that 50% of all assessments on every course unit must be AI-proof. We could, then, just tick off as ‘sorted’ the risk of students using Generative AI to complete their coursework assignments for them. Some universities, indeed, have already gone down this sort of route quite heavily.

One thing that such a scenario would also entail, though, is hugely increased pressure on staff. Academic staff would simply have to get on with life as best they could. Higher Education, though, really is a very stressful profession to work in if it's all down to you and your own individual practice. After all, we prioritise research groups in universities, recognising that high-quality practice occurs when practice is shared.  

Addressing the challenges posed by AI is no different, and any response that the University looks to develop will benefit hugely from colleagues pulling together. Along with some colleagues, I recently conducted a study that looked at the response of a group of staff and students to use of GenAI in assessed work. One of the conclusions from the study, which you can access in the journal Learning, Media and Technology, was that there is a value in a collective response on the part of teaching staff.

For instance, there is plenty of uncertainty for many staff where Generative AI is concerned, not least because many students have greater knowledge of various AI tools than we do. I am sure, for instance, that some of my students on the MA Digital Technologies, Communication and Education will have been playing around with AI agents that can complete online tasks. By contrast, though, I have not yet used an agentic browser (at least at the time of writing this blog), let alone used one to complete substantive tasks that I would otherwise have had to perform myself. If staff can find ways to share with each other understandings of how to respond, though, then such uncertainties will be ameliorated, and constructive action will be much more likely to occur than if we are just all left to our own devices. 

Given all of this, I have been working within the School of Environment, Education and Development to develop our infrastructure for shared practice across the School. We need platforms, shared routines, collectives, online resources, funded projects, examples of agreed ways forward and so on that make it realistically possible both to develop more secure forms of assessment and to foster appropriate student expertise for GenAI.  Some ideas are outlined in the SEED strategy for Generative AI in Teaching and Assessment, with a set of associated resources available alongside the strategy. The AI Commons initiative that hosts this blog posting also very much fits into this space.

🎤 FBMH GenAI in Teaching Seminar Series

The Faculty of Biology, Medicine and Health AI-Teaching & Learning Group is running a seminar series for staff using generative AI in teaching. Each session features two short talks (5–15 minutes each) followed by Q&A, and runs both in person and on Teams.

If you'd like to present or attend, contact [email protected]. To join the mailing list for future seminars, email [email protected].

💭 Something to think about

Normalising AI through institutional provision ultimately reshapes expectations in ways that have significant implications. At the same time there is accumulating evidence that students feel pressure to use AI not because they value it but because they fear falling behind those who do. Choosing not to use it, for pedagogical or ethical reasons, risks looking like resistance rather than principle. In such conditions, the most intellectually autonomous students may be quietly disadvantaged while more strategically minded students may benefit, and as a consequence the overriding climate becomes one of suspicion.

From Free AI or AI Free? by David Spendlove

👋 First impressions of Copilot 365?

Professional services colleagues now have access to Copilot 365, and academic staff will follow soon. Many colleagues have already been using it through the early release cohort.

We'd love to hear how it's going. What's working? What isn't? What took you by surprise? Whether you've been using it for months or have just started, your experience would help others know what to expect.

Get in touch: [email protected]

We hope you enjoyed this fifth issue of the AI Commons. If you found it valuable, would you consider forwarding this newsletter to your colleagues? Comments, suggestions and questions are always welcome.