4.2. Future work
Beyond ChatGPT, other GenAI tools and models such as Midjourney, Stable
Diffusion and DALL-E 2 that automatically generate videos, and images
within minutes; Eleven Labs that transforms text into audio; and Magic
Write that assist creative writing (Peres et al., 2023), can potentially
influence students’ everyday use of AI to assist their coursework. The
full potential of GenAI in general and specifically on education is
still unknown and underpins a need to investigate its impact in a manner
that considers students’ points of view (Gašević et al., 2023) and their
development. Accordingly, efforts to better align and assess the
effectiveness of the learning agreement from a self-determination theory
perspective should be explored. For example, the UNESCO guide providing
an overview of how ChatGPT works and can be used in higher education,
highlights the importance of autonomy in suggesting that guidance on
“how and when ChatGPT can be used (and when it cannot) … should be
negotiated with students and teachers, not imposed on them” (Sabzalieva
& Valentini, 2023; p. 13). Investigating effects on student feelings of
autonomy in comparison to other ways of communicating academic integrity
and course policies can be explored. The same UNESCO guide suggests that
GenAI should relate to course learning outcomes to help students
understand how it “can support their learning and what expectations
there are for them” (Sabzalieva & Valentini, 2023; p. 13). Thereby,
helping students develop a competence around figuring out how it can
best serve their needs and around what uses align with their value
systems. Future work can investigate the effectiveness of different
learning agreement designs and implementations in terms of impact on
student decision making – and pair the use of learning agreements with
approaches to help students make more thoughtful decisions (Arvai et
al., 2004) related to GenAI use. This support of thoughtful decision
making can also explore and assess how effective learning agreements are
as tools for raising awareness of ethical concerns as identified by
researchers, policy makers, and the public. Concerns such as
discrimination – and the need for transparency and accountability
(Nguyen et al., 2023; Akgun & Greenhow, 2022; Miao & Holmes, 2021); an
increasing digital divide – accessibility limits due to internet
availability and/or regional access restrictions (Sabzalieva &
Valentini, 2023; Miao & Holmes, 2021); concentration of power and abuse
of power (Sabzalieva & Valentini, 2023); surveillance and intrusiveness
(Miao & Holmes, 2021; Akgun & Greenhow, 2022); loss of autonomy
(Nguyen et al., 2023; Akgun & Greenhow, 2022); obstructing holistic
competencies (Chan & Hu, 2023); and lack of ecological sustainability
(Nguyen et al., 2023). In relation to the last aspect of
self-determination theory, relatedness, future work can compare learning
agreement construction and use at individual and group levels to
investigate the impact of relatedness on student decision making and
understanding of the ethical concerns related to GenAI use. Finally, the
design and implementation of the learning agreement can be improved
following the suggestions made by students which align with research
related to improving the effectiveness of another type of agreement,
informed consent forms. Improving readability, including causal
explanations, and supporting understanding with enhanced discussions can
help overcome cases in which student prior knowledge of the items in the
learning agreement are impoverished (Beardsley et al., 2019).