No life, no understanding: the limitations of GenAI as an interpretative interlocutor

This paper argues that outsourcing the interpretation of human experience of phenomenon to GenAI risks missing what matters to people, what an experience feels like for them and ultimately missing the human point entirely. Written as a first-hand account it follows a novice Interpretative Phenomenological Analysis (IPA) researcher experimenting with Gemini AI and CoPilot as interpretative research assistants in the analysis of qualitative interview data on student experience. Structured as a three-act screenplay, Act 1 recounts an uncomfortable discovery: both GenAI agents were drawing on outdated IPA methodology. Confidently wrong. Act 2 concerns the philosophical heart of the matter. Can a GenAI agent without lived experience, fore-structure, or reflexivity, engage meaningfully with phenomenological interpretation? Act 3 offers hope. The GenAI agents proved genuinely useful as a productivity co-design buddy in processes surrounding interpretation: audit trails, quality frameworks and rapid prototyping. The conclusion: the interpretative heavy lifting remains stubbornly, necessarily human.