Negotiating agency: university students’ experiences of co-creating content with generative AI
Аннотация
While generative artificial intelligence (GenAI) is widely utilized for content creation, the ongoing debate regarding the interplay between human agency and GenAI remains unsettled. Grounded in Bandura’s (2001) theory of agency, this qualitative study explored a cohort of college students’ experiences with GenAI (specifically ChatGPT) for content creation by investigating the determinants and influences on the use of GenAI, as well as their agency in the GenAI-assisted content creation process. The study employed semi-structured in-depth interviews with fourteen university students, coupled with the collection of screenshot records of their GenAI activities. Drawing on thematic analyses, the findings suggested three primary factors influencing student GenAI adoption, including efficiency, high-quality content, and inspiration/ideation. Throughout the content creation process, students excised their agency as they engaged in critical thinking while using GenAI, actively thinking about, evaluating, and adapting their content creation behaviour. This study also showed that the use of GenAI yielded mixed effects on students’ decision-making ability and creativity. It emphasized the changes in students’ decision-making roles in content creation and the sense of “restriction” they experienced. These findings offer insights into how to balance students’ human agency with their use of GenAI, particularly by highlighting the importance of transparency in GenAI systems and emphasizing the need to foster students’ critical thinking throughout the GenAI-assisted content creation process.