AI literacy as a threshold concept. Fostering sustainable AI education

Abstract

The rapid developments in the field of artificial intelligence (AI) and Large Language Models (LLMs), in particular, clearly indicate the growing need for the teaching of basic AI competencies (AI literacy) as part of a sound and sustainable university education embracing digital transfor-mation (United Nations Sustainable Development, 2023). According to Long and Magerko (2020, p. 2), AI literacy is defined as “a set of competencies that enables individuals to critically evalu-ate AI technologies; communicate and collaborate effectively with AI; and use AI as a tool online, at home, and in the workplace.” Among others, these competencies include the ability to recog-nize AI-based tools, a fundamental understanding of how artificial intelligence operates, an un-derstanding of the strengths and weaknesses of AI, a reflection of the human role and ethics regarding AI, and an understanding of the programmability of AI agents (Long & Magerko, 2020, pp. 4-8). Identified as one of the 21st Century Skills (Casal-Otero et al., 2023; Ng et al., 2023; OECD, 2019; Southworth et al., 2023; Van Laar et al., 2017, p. 578; World Economic Forum (WEF), 2018, p. 37), students need to learn how to use AI tools in a critically reflective manner to become competent “managers of AI,” as Cardon et al. (2023, p. 277) explained. In their view, AI literacy emphasizes the practical and responsible application of AI tools beyond a basic com-prehension of digital abilities. This should also include addressing ethical concerns in the imple-mentation of AI in education, emphasizing transparency, bias mitigation, data protection, collabo-ration, and accountability to ensure ethical AI use (D’Souza et al., 2024, p. 1-2) as well as the critical reflection of the impact that AI technologies have with regard to environmental sustainabil-ity, such as their ecological footprint caused by high energy consumption and greater emissions (Floridi, 2024, p. 7-8). Since AI literacy has significantly gained in importance and is likely to con-tinue to be of great relevance, it might be reasonable to regard it as a key concept in higher edu-cation. In fact, it might be conducive to define it as a “threshold concept,” following Meyer and Land (2003). According to them, A threshold concept can be considered as akin to a portal, opening up a new and previously inaccessible way of thinking about something. It represents a transformed way of understanding, or interpreting, or viewing something without which the learn-er cannot progress. As a consequence of comprehending a threshold concept there may thus be a transformed internal view of subject matter, subject landscape, or even world view. (Meyer & Land, 2003, p. 1) In other words, a threshold concept can be viewed as a “transitional point or intersection,” helping the learners navigate their way through a “transformational landscape” – such as the continuously changing AI landscape (Meyer & Land, 2005, p. 379). Due to these characteristics, threshold concepts have become quite popular and accepted in educational the-ory (Nicola-Richmond, 2018, p. 101). Inspired by Meyer and Land’s work, we thus propose con-ceptualizing AI literacy as a threshold concept in the broader sense and to integrate AI literacy into higher education teaching as a “key point of curricula” (Kilgour et al., 2019, p. 1417). We believe that this approach is highly promising to foster this key 21st Century Skill as part of a socially sustainable, future-proof education and to thereby transform higher education long-term. In this conference paper, we would like to explore how this approach could be realized, based on the findings from a research project which we conducted in the academic year of 2023/2024. With this, we hope to make a relevant contribution to the field of sustainable AI education.