Since Generative AI really made its presence felt in higher education in early 2023, the Faculty of Health have always been up for an opportunity to connect and share GenAI experiences from within and beyond their faculty. After workshops on the topic in May and November in 2023, a project from an interdisciplinary team across disciplines was established last year through a Learning and Teaching Scholarship in the Faculty of Health, led by Prof Bronwyn Hemsley: UTS GenAI-ENHANCE (Enabling Networks of Health professionals to Create and Evaluate opportunities in GenAI). A key objective of this project was to create a Community of Practice around GenAI and Health, and convene a summit on GenAI exploring the many advances across disciplines in both teaching and research.
As attendees settled into the Aerial auditorium for the day it was quickly established that, while the fast pace of GenAI remains a disruptor, it should no longer be the dominating focus if we are to move forward. A/Prof Lynn Sinclair (Associate Dean, Teaching And Learning, Health) saw our current standing primarily as an opportunity to think differently about curriculum. Prof Kylie Readman (DVC Education And Students) noted that we are beyond trying to control or subdue GenAI, as it is now no longer appropriate to ban the use of GenAI in subjects. She put forward that we need to be teaching ourselves and leading the way through the change, while talking and listening to (and learning from) our students. And A/Prof Jan McLean outlined how UTS has responded to calls for strategic development and implementation of GenAI.
Developing AI literacy
We continue to develop and update resources at UTS to guide our community on how to keep up with a fast-changing topic, and it’s useful to hear how others are developing and sharing guides to develop AI literacy. Dr Ananthan Ambikairajah (University of Canberra) surveyed students to assess GenAI usage trends, the insights of which instigated online workshops to understand AI tools ‘under the hood’, compare GenAI action plans from a range of universities and share prompt engineering examples. Dr Nici Sweaney from AI Her Way suggested that, in getting started with GenAI, we should focus on LLMs, build our AI literacy gradually and focus on a small number of essential tools when starting out.
ARC Future Fellow A/Prof Heather Ford advocated for being actively involved in conversations about how GenAI develops. She identified one of the biggest threats to AI as people’s passive or unquestioning acceptance of claims made by this ‘confidently wrong’ instructor.
Listening and learning
Analytics specialists from UTS reminded us that we need to always be thinking about the humans behind the data. Prof Simon Buckingham Shum noted our planets’ disturbing climate trajectories and how/where/if environmentally damaging tech fits into this landscape. His provocations for staying orientated amongst the noise was to listen to: our planet, our students, the research and our prototypes. Susan Gibson, head of data analytics and UTS, followed this up by introducing the UTS Generative AI framework, where you can load content into a folder, create prompts and enable it for students. This will empower users to create their own AI assistants over existing data and systems, and it includes chat history – contact Susan via email for more details on the framework.
A focus on Health
Throughout the day there were some fascinating insights into how GenAI was being used and integrated in health areas:.
- Prof Amanda Wilson began experimenting with AI in 2022 and started incorporating it into her workflows to refine questions, speed up processes and offer alternative insights. Devising and refining prompts for evaluation can be time-consuming but once AI content is generated, it saves time and makes for a more seamless experience in the long run. She suggested getting familiar with a variety of GenAI tools before using CoPilot as the safe, secure tool for UTS work.
- Dr Julia Dray is engaged in a small-scale study funded by the UTS Disability Research Network codesigning GenAI solutions for child and youth mental health. The more personalised therapy and tailored interventions that AI offers in this domain could create engaging, therapeutic content with improved accessibility and reach.
- Dr Lucy Bryant’s team explores using VR and AR in the field of communication disability. She explained how GenAI-responsive avatars allow for a more interactive style of storytelling, but it’s a multi-step process that causes a lag and can lessen the immersion. With AI allowing for a higher level of personalisation of environments and avatars, she noted the ethical considerations of being exposed to an increasingly realistic mixed reality environments.
- Danielle Gardner was awarded a Faculty of Health L&T Scholarship in 2024, to pilot an AI-generated communication training platform, SimConverse, for supporting students in Nursing and Midwifery to practice taking a patient’s case history at the bedside. Improvements since the initial run include a microphone system for clearer communications and better converse from different locations, and foot pedals to activate the AI chatbot from different locations.
Innovations in external tools
AI avatars continued to make their presence felt throughout the day. In his role as a simulation specialist at Evolve Simulations, Soren Jensen inhabits bespoke computer avatars as form of puppeteering. He related to GenAI in its interpreter/creator role and discovered that the highly flexible AI avatars could encapsulate a wider range of characters more confidently. Soren outlined emerging outcomes of a collaboration with speech pathology at UTS, aiming to provide a way for students to practice responding in objective structured clinical exams.
Co-designed with children, the speech therapy application Say66 engages children, accelerates learning and normalises therapy via relatable gaming-style activities. Say66’s Prof Kirrie Ballard explained how they are now utilising AI agents, with plans to embed it into the game as multiple agents. This innovation empowers families to get access to something they haven’t had before and saves time in sourcing materials, lessening the cost/time pain points that impact on therapy provision across Australia.
Ethics in AI
A common point of the day’s presentations from UTS staff was considering the ethics of GenAI, with concerns ranging from privacy to inequality.
UTS research assistant and qualified lawyer, Fiona Given shared her lived experience of the challenges of using augmentative and alternative communication speech generating devices, such as keeping pace with conversations and the physical effort of entering each word into the device. GenAI can offer improvements such as better speech recognition and whole paragraph text generation, but incorporating it also comes with uncertainties and risks. People with disability using the GenAI in a device would need to have full control over what was spoken, and therefore need a way to ‘check and correct’ prior to the device being activated to speak the message. How will we know if the person is talking? And will communication device users have to disclose that their message was generated with the assistance of AI?
With the ethical considerations always in play and tool advancements coming faster than many of us can keep up with, learning through the shared experiences of events such as today’s summit are as important as ever.
To find out more about the GenAI-ENHANCE project, including the Community of Practice, contact the project lead Bronwyn Hemsley.
