Your goal: show AI product judgment: define what "wrong" means and measure it, choose the fix
that fits the cost, then make a keep-or-kill call with evidence and a decision date.
How this works (about 43 minutes)
- Round 1, AI product sense (about 22 minutes): a trust problem with an AI assistant, with follow-ups and a cost curveball.
- Round 2, strategy (about 20 minutes): a product that has not found traction. Ask Ben for any numbers you need; they are made up for the exercise.
- Ben tells you when he switches. This is a practice room in the style of an AI-lab loop, built from what candidates report; formats differ by lab.
- If you want to jot a structure, ask Ben for the notepad and tell him when you have written something.
- No hints and no scores during the call. Your report comes after.
What scores well
- Saying what "wrong" means, who is hurt most, and how you would measure it
- Fixes matched to cost and speed, not "add citations everywhere"
- A decision rule for keep or kill: the bar, the date, and what would change your mind
- Pricing a "strategic" claim instead of accepting or dismissing it
Tips
- Take the AI product sense and strategy coaches first.
- Name the tradeoff out loud: quality, cost, speed, trust.