As we approach the end of 2023, the Faculty of Health Teaching and Learning Team and partners from IML and the LX.lab held a workshop on generative AI in learning and teaching to reflect on the year and explore ways forward.
The workshop included some evolving ways to think about generative AI from Ann Wilson and Caroline Havery (IML), and a presentation on prompt priming from Dr Suman Laudari (Charles Darwin University). After getting interactive with various GenAI tools, Faculty of Health colleagues presented updates on their research, projects and assessment designs in a showcase which built on their Generative AI in Learning and Teaching workshop from earlier this year. With questions and discussion flowing right through to drinks, there was plenty of appetite to keep learning more.
GenAI: Red, Amber, or Green for Go?
Ann Wilson and A/Prof Caroline Havery kicked things off with an overview of a ‘traffic light’ system for thinking about how generative AI is used. They mapped usage to TEQSA‘s recently published guiding principles and propositions in Assessment reform for the age of artificial intelligence, suggesting three tiers of application:
- 🟢 ‘Green light’: GenAI is integral – use freely
- 🟡 ‘Amber light’: GenAI as an assistant
- 🔴 ‘Red light’: GenAI is NOT to be used
Read more about this approach in a new series of blogs by Ann Wilson, Caroline Havery, Rosalie Goldsmith and Jenny Wallace: Assessment in the age of GenAI: suggestions for a way forward
Getting hands-on with prompt priming
Digital Learning Designer and former UTS staff member Suman Laudari kicked off the next session with some shared insights and tips from his Times Higher Education resource series on generative AI in a higher education context. From simple starts to more complex, nuanced and directive prompting, Suman outlined how careful prompting can refine rough drafts into increasingly useful outputs.
Armed with new ideas and inspiration, workshop attendees rolled up their sleeves, opened up their laptops and got stuck into some prompt priming, with careful instructions and facilitation from Jenny Wallace and Lucian Sutevski. The starter prompt was:
Suggest three ways of persuading people who are vaccine hesitant to get vaccinated. For each suggestion, discuss the pros and cons. Provide evidence for each suggestion.
Using any of ChatGPT, Bing, Bard or other GenAI tools, the task was to keep refining the prompt in different ways to generate a desired output. Collated in a shared google doc, the results showed a vast range of different approaches and outputs, from formal reports to interactive role plays, generating much discussion on the impacts and potential use of this new set of skills.
Showcase: what have we learned with GenAI so far?
The final session of the afternoon showcased 4 presentations from academics across the faculty who have been running research, projects and assessment designs on the use of generative AI, sharing continued learnings since the Generative AI in Learning and Teaching workshop from earlier this year. Presenters were:
- Dr Daniel Demant (School of Public Health) – results from research exploring student perspectives on GenAI, and whether this is impacted by psychosocial factors
- Prof Amanda Wilson (School of Nursing and Midwifery) – who shared findings from 3 separate projects, including a Midwifery assignment where students used GenAI for a background paper, critiquing and analysing outputs before creating an informative digital artefact
- Dr Mike Rennie (School of Sport, Exercise and Rehabilitation) – using GenAI in a High Performance Science assessment, to evaluate interventions and compare recommendations with ChatGPT (also see this AI Case Study resource)
- Prof Bronwyn Hemsley (Graduate School of Health) – shared a range of projects underway in Speech Pathology, focussing on aspects such as GenAI’s use in special education, interactions with social media, and more.
You can also read about the examples from Sports Science and Speech Pathology in Assessment in the age of GenAI: GenAI as an assistant.
Bronwyn concluded by noting that activity in this space is more cautious than adventurous at the moment, warning that “it could become a danger in health circles if you don’t know how to use it”. By the end of the workshop, the mood was curious and optimistic, with no shortage of ideas for ways that GenAI could be explored further in the coming year.