Prompting Higher Education Towards AI-Augmented Teaching and Learning Practice

Journal of University Teaching & Learning Practice Volume 20 Issue 5 Quarterly Issue 2 Article 02 2023 Prompting Higher Education Towards AI-Augmented Teaching and Learning Practice Bronwyn Eager University of Tasmania, Australia, Ryan Brunton University of Tasmania, Australia, Follow this and additional works at: https://ro.uow.edu.au/jutlp Recommended Citation Eager, B., & Brunton, R. (2023). Prompting Higher Education Towards AI-Augmented Teaching and Learning Practice. Journal of University Teaching & Learning Practice, 20(5). https://doi.org/10.53761/ 1.20.5.02 Research Online is the open access institutional repository for the University of Wollongong. For further information contact the UOW Library: Prompting Higher Education Towards AI-Augmented Teaching and Learning Practice Abstract Large Language Models (LLMs) and conversational-style generative artificial intelligence (AI) are causing major disruption to higher education pedagogy. The emergence of tools like ChatGPT has raised concerns about plagiarism detection but also presents opportunities for educators to leverage AI to build supportive learning environments. In this commentary, we explore the potential of AI-augmented teaching and learning practice in higher education, discussing both the productive affordances and challenges associated with these technologies. We offer instructional advice for writing instructional text to guide the generation of quality outputs from AI models, as well as a case study to illustrate using AI for assessment design. Ultimately, we suggest that AI should be seen as one tool among many that can be used to enhance teaching and learning outcomes in higher education. Practitioner Notes 1. Learning to write effective instructional prompts for AI models will help augment learning and teaching practice. 2. AI models offer the potential for significant productive affordances, including personalised feedback, adaptive learning pathways, and enhanced student engagement. 3. To successfully integrate AI into higher education, institutions must prioritise faculty development programs that provide training and support for educators to effectively use these technologies in the classroom. 4. Institutions must ensure that AI is used in a way that aligns with their values and mission and that students are informed about how their data is being used. 5. It is important to recognise that AI is not a panacea for all of the challenges facing higher education. Rather, it should be seen as one tool among many that can be used to enhance teaching and learning outcomes. Keywords ChatGPT, artificial intelligence, large language model, assessment design, prompt engineering This commentary is available in Journal of University Teaching & Learning Practice: https://ro.uow.edu.au/jutlp/ vol20/iss5/02 Eager and Brunton: Prompting Higher Education Towards AI Introduction The higher education community has been galvanised by the mainstream emergence of Artificial Intelligence (AI). Worldwide, universities are grappling with the broader implications of these technologies (Bjork, 2023), with AI threatening traditional assessment design including the demise of essays and online assessment (Cassidy, 2023). As global awareness of AI snowballs, and the availability and adoption rates of consumer-based AI tools skyrocket, universities are increasingly recognising the need for teaching and learning approaches to evolve in a manner that adapts to a shifting landscape shaped by the growing influence of AI. However, opinions on what that path forward will look like remain far from clear. Practical responses to 'the AI situation' vary widely. On one hand, people may resist AI or deploy defensive measures. Our observations suggest that such responses originate from an initial awareness of AI tools, leading to curiosity, followed by experimentation, and finally, resistance. This trajectory may result in attempts to regulate AI usage, including outright bans on AI technologies (McCallum, 2023) and the adoption of AI-detection software. Resistance likely has origins in concerns that AI tools can produce essays, answer questions to exams, and augment student capability to achieve an advantage when completing any form of assessment task involving computer-based work (Cotton et al., 2023; Kung et al., 2023). We recognise these concerns and acknowledge AI's potential to be used in ways that displace or misrepresent human effort. However, we believe in taking a different lens to the AI debate. We advocate for the integration of AI technologies across academia as a way of potentially improving teaching and learning practices, while ensuring the continued relevance and sustainability of higher education. In this commentary, we draw from our roles as an academic educator and a learning developer to extend earlier discussions in this Journal. We build upon Perkins’ (2023) exploration of AI's role in academic integrity and Crawford et al.’s (2023) inquiry into the ethical use of AI models. By leveraging and sharing our experiences, we seek to assist those who are aiming to create highquality AI outputs and navigate this transformative era. In what follows, we first discuss the importance of learning how to write effective instructional commands to optimise the usefulness of AI tools when generating teaching and learning content. The discussion is framed by an introduction to ‘prompt engineering’ and suggestions are provided for different approaches to writing effective prompts. We then proceed to operationalise these insights by presenting a case study with which to showcase how prompts can lead to augmenting assessment design. We conclude by noting practical considerations for guiding the high education sector towards a sustained future amidst the predictable state of operating within an AI-ubiquitous world. Academic Editors Section: Curriculum and Assessment Editor in Chief: Dr Joseph Crawford Senior Editor: A/Prof Michael Cowling Publication Received: 3 May 2023 Revision: 21 May 2023 Accepted: 23 May 2023 Published: 29 May 2023 Copyright: © by the authors, in its year of first publication. This publication is an open access publication under the Creative Commons Attribution CC BY-ND 4.0 license. 1 Journal of University Teaching & Learning Practice, Vol. 20 [2023], Iss. 5, Art. 02 Prompt Engineering Artificial Intelligence (AI) involves machines simulating human intelligence processes. One type of AI, known as Large Language Models (LLM), can produce text that resembles human writing. A specific application of a LLM is a ‘chatbot’, perhaps the most well-known being ChatGPT. However, there are many chatbots, including Google’s Bard and Microsoft’s Bing. These software programs work by mimicking human conversation, interacting with users primarily through text: upon receiving a text-based input, chatbots analyse the patterns in the data they've been trained on, then use these patterns to predict and g (...truncated)