Fifty voices, before the apero. Findings, next morning.

A global consumer brand had three possible futures on the table on the way to a long-horizon product strategy. A citywide optimisation layer for urban living. An assistant that anticipates daily decisions. A life-long personal health system. Each raised hard questions, about autonomy, surveillance, and fairness. Naming the futures was the easy part. The hard part was getting fifty senior leaders, the bench two and three levels down, to engage with them honestly and in depth, before the session moved on.

Each leader sat one to one with an adaptive AI interview agent, in the language they think in. The agent asked for specifics where an answer stayed general, and pressed on the lines people were drawing. Fifty conversations ran in parallel and finished before the apero. Overnight, the synthesis read each scenario back: what drew people in, what worried them, the lines they would not cross, and where they genuinely disagreed.

A vote would have collapsed the room into one number. The synthesis did the opposite. On a single scenario, six distinct closing positions held, each in the leaders' own words, from enthusiastic adopters to ethical observers watching for skill erosion. One line ran under all of them: trust in data-driven optimisation, but only as far as individual or civic voice was respected. Convenience was welcomed. Manipulation was refused.

The leadership team did not open the next workshop block cold. They opened it seeing where the room actually stood, in the room's own words, and used that as the input to the rest of the day. A written record of senior thinking on three loaded futures stayed with the company as the real product bets were placed.