Discernment learning labs: architecting insight-generating conversations

As generative AI makes plausible outputs easier to produce, a central design challenge is no longer only how to generate options, but how to organise inquiry so participants convert such outputs into warranted claims and defensible commitments. This paper reports a design-science instantiation of an undergraduate financial management course architected as a Discernment Learning Lab: a structured setting for insight-generating conversations in complex and contested situations. Students worked in project teams and same-role communities of practice across interpreting and creating contexts, using a shared artefact set comprising a common-orientation framework, role-based prompt scaffolds, insight-generating operators, and a dual evaluation architecture, under open, integrating, and edifying norms of participation. Discernment effects were illuminated through Learning Tables, Ethics Tables, and reflective essays, performance effects were evaluated through key performance, relationship, and situation indicators (KPI/KRI/KSI). The instantiation shows that AI can accelerate representational and exploratory work without displacing human responsibility for judgement, commitment, and action.