Answers are grounded in this paper and cite their sources.
Postgraduate supervisors acquire their expertise through practice, not formal training, and rarely articulate it. Research on AI in higher education has focused on students and assessment, not on the supervisors themselves. This study asks what supervisors believe supervision is for, and how that belief surfaces when AI enters the conversation. Semi-structured interviews with 22 capstone supervisors were analysed through reflexive thematic analysis. Rather than evaluating the technology, they used the AI question to articulate, often for the first time, the beliefs underpinning their work. We term these beliefs the pedagogical credo, expressed as three stances, each a distinct understanding of what makes supervision irreplaceable. We propose the looking glass effect for how that discussion gave concentrated form to beliefs previously diffuse in practice. The findings suggest institutions should create occasions for supervisors to articulate what guides their work, before it is lost to a question no one asked.
No takes yet. Share an insight, caveat, or question.
Levenberg et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: