Product design question

How would you design a bridge connecting Africa with South America?

Practice this question out loud. An AI interviewer asks it, follows up like a real interviewer would, and scores your answer. Type or speak.

What this question tests

Handling a deliberately absurd or infeasible prompt: can the candidate reframe it sensibly rather than attempting a literal answer.

Recognize the prompt is likely testing reasoning and structured thinking under an absurd constraint, not literal engineering feasibility, and say so.
Reframe with the interviewer: clarify whether they want a literal engineering approach, or the underlying business case for connecting these continents.
If treated literally, name the fundamental infeasibility: the distance across the Atlantic at the narrowest point is still thousands of kilometers, far beyond any existing bridge technology.
Pivot to the underlying need: what problem would connecting Africa and South America actually solve, likely trade and travel time, and propose alternatives like expanded shipping or flight infrastructure instead.
If the interviewer insists on a literal answer, decompose the request into a staged engineering approach: research existing longest bridges and tunnels for comparison, identify the shortest viable crossing point, and estimate order of magnitude cost and timeline, clearly flagging this as far beyond current engineering capability.

Explicitly calls out the prompt's impracticality up front instead of pretending a literal bridge is feasible, showing calibrated judgment.
Reframes toward the real underlying need, trade and travel connectivity, which is what a good PM would actually solve for.
If pushed for a literal answer, grounds it in real comparisons, like the world's longest existing bridges, rather than inventing numbers with no anchor.

Answering as if a literal transatlantic bridge is a realistic engineering project without acknowledging its infeasibility.
Failing to check with the interviewer what they actually want to learn from this unusual prompt.

What underlying business problem do you think this question is really getting at?
If a bridge is infeasible, what alternative would you propose instead?