Ask generative AI a question, and you get a confident answer. What you don't get is a second read, a dissenting take, a reason to dig deeper. When open educators use AI to create and remix resources that others will reuse, that untested confidence travels with the work.
The counterintuitive fix? Sow discord. Not by asking AI to see both sides, but by designing personas with named stakes that structurally conflict, then putting them in the same room. When multiple perspectives are built for friction, disagreement surfaces ideas neither would reach alone. Then run the output back through a critical lens, round after round, until what remains has been earned by both sides.
Open educators make decisions alone that deserve a panel, and this roundtable method convenes one on demand. Attendees leave with a replicable roadmap for building and running their own expert panel, no subject matter expertise required.
Attendees of this session will be able to:- Explain why single-prompt AI use produces untested confidence and why that matters when the work will be reused by others.
- Apply the roundtable method by assigning personas with structurally conflicting stakes to surface complex ideas no single perspective could reach alone.
- Use iterative review loops to pressure-test AI output and recognize what repeated scrutiny reveals that a single pass misses.
