Placements, and the associated discussions with supervisors and mentors, co-workers, patients and caregivers, provide valuable learning opportunities for our students. But with increasing pressures on the NHS and related organisations, time with mentors and supervisors can be hard to come by – and students don’t always get a chance to practice some of the skills they will need to thrive post-graduation. Can the recent advances in generative AI help fill some of the gaps and complement in-person learning experiences?
The Digital Education Community of Practice (DECoP) May webinar addressed that question. Dr Jo Blannin, Senior Lecturer, Digital Transformations and Deputy Director of the Hub for Educational Design and Innovation, at Monash University in Melbourne joined us to share her work using generative AI to create personalised and authentic learning simulations.
Jo shared how the ATLAS (Authentic Teaching and Learning Application Simulations) platform was developed by an academic in Public Health in response to a need for students to be able to practice communicating public health messages on social media. Dr Joel Moore created a set of AI-powered social media personas, which responded to and shared students’ posts within a closed environment, allowing students to see how people might respond to particular messages and framings and how these messages might get twisted over time.
More simulations from a range of disciplines have since been added to the platform. Students studying Teaching, for example, can practice discussions with parents about their child’s education – an important skill, but one that students do not get to practice while on placement. Similarly, Nutrition and Dietetics students can practice conversations with patients, test out different approaches to these conversations and learn how to ‘rescue’ a conversation that has taken a negative turn.
While generative AI can be used to ‘power’ the simulations, Jo highlighted the importance of the teaching team and other stakeholders, including placement partners and patients or clients, developing realistic personas themselves. Using AI to generate personas is likely to result in ‘reversion to the mean’, and not represent the diversity and complexity inherent in our communities and the people who whom we work.
Jo also stressed the ATLAS simulations – and other simulated learning activities – should be used in conjunction with, and to supplement, ‘real’ practical and placement activities, and not as a replacement.
Watch the webinar below.
The next DECoP webinar will be presented by Mr Kash Aktar, Reader in Orthopaedic Surgery and Surgical Education in the Blizard Institute and co-host of the successful Orthohub podcast. Entitled "From Pixels to Personalities: humanising digital learning through storytelling", the webinar will be held on Thursday 11 July at 11am, via Teams. Join the DECoP Teams site for updates and the meeting link.