This paper explores the use of AI-generated "Shadow Podcasts" as a supplementary learning tool in higher education. Developed using Google's NotebookLM, the podcasts transformed lecture transcripts into short audio summaries, offering students an alternative means of engaging with course content. Deployed across five modules in two academic Schools within the Robert Gordon University, the podcasts were evaluated through a mixed-method survey of 85 students, combining quantitative ratings with thematic analysis of open-ended responses. Analysis suggests the podcasts were well-received, with most students rating them as good or excellent and noting improvements in their own engagement, understanding, and revision. The tool's conversational tone and ability to support multitasking distinguished it from traditional materials. Critiques centred on the artificial delivery, and lack of visual or detailed content. Students expressed interest in improvements such as more expressive voices, subtitles, video integration, and stronger alignment with assessments. The study highlights the potential of generative AI to enhance educational experiences when used transparently and reflectively. While Shadow Podcasts are not being positioned as a replacement for live teaching, they offer a flexible, scalable, and engaging complement to existing resources. Future work will involve cross-institutional research and further exploration of the podcasts' impact on learning outcomes and curriculum design. This study contributes to the growing dialogue on pedagogically-grounded uses of generative AI in computing and business education.
Pirie et al. (Tue,) studied this question.