I asked an AI to interview me about my own biases before I wrote a single research question.
That was the exercise: run a structured prompt end to end on a real product, and see how far it gets you on day one versus what it would normally take months to learn.
I picked Xbox Game Pass. The prompt asked me one question first — what’s the product — then played back its own read of the target customer and the user journey, and asked me to correct anything it got wrong. Only after that did it draft ten research questions.
The interesting part wasn’t the questions. It was watching it catch its own mistakes. It flagged “what makes playing through Game Pass difficult?” as an assumption that difficulty exists at all. It flagged “why did you subscribe?” as a question that invites a rationalized answer instead of the real decision. Then it rewrote both — “tell me about a time Game Pass didn’t have something you wanted, what did you do next” instead of assuming friction, “think back to when you first subscribed, what was happening at that time” instead of asking for a reason.
That’s the actual skill this session was teaching. Not writing questions — noticing the ones that already contain their own answer.
The instructor’s point about this stuck with me: “I would have never thought of these questions if I was going in as a fresher. It would take six or seven months to get here.” AI doesn’t replace the judgment. It gets you a competent first draft before you’ve earned the judgment yourself.
#ProductManagement #AIProductManagement #ArtificialIntelligence #ProductThinking #ProductManager #Masai #IITRoorkee #LearningInPublic #ContinuousLearning #CareerGrowth #AIPM