AI doesn’t always behave the way we expect, and that’s the point! In this joint talk, one speaker plays the skeptic and the other the advocate as we explore how the randomness of AI can actually work in our favor.
- When AI produces a weak or confusing answer, it’s a teachable moment: how can we reframe the prompt to get something better?
- When AI exceeds expectations, it shows the power of refining and building on strong outputs.

Through live demos, playful role reversal, and audience interaction, we’ll show that whether AI delivers worse than expected or better than expected, developers can always spin it into actionable insight. Attendees will leave with practical strategies for iterating on prompts, handling unpredictability, and making AI a reliable partner in their workflow.