AI could make getting people in a room more valuable

I’ve spent a substantial part of my product career getting people into rooms to work through problems. Design sprints gave that work a structure: frame the question, explore possible answers, make something tangible and decide what to test.

As AI takes on more of the production involved, I’ve found myself wondering which parts of that process were actually doing the useful work.

If we can arrive with several plausible directions and something people can try, there may be less reason to spend a day generating and arranging ideas. There is still a reason to discover that the people around the table mean different things by “better”, or that the preferred solution asks one team to absorb costs that another team hasn’t considered.

A prototype can make a proposal easier to understand. It doesn’t establish that the people expected to deliver it agree with the assumptions behind it.

That makes me think the shared time could become more valuable, even as we need less of it. We could spend more of it examining trade-offs and deciding what we are prepared to commit to.

The counterargument is that this gives meetings too much credit. People can appear to agree because a senior person has spoken or because everyone wants to leave. A team with clear authority to try something may learn more from a small experiment than from another discussion. In a fast-changing environment, too much emphasis on agreement can make changing direction unnecessarily expensive.

I think the distinction is how many people need to act differently for the decision to work. A reversible product experiment may need very little collective deliberation. A change to how several teams serve customers probably needs considerably more.

I’m increasingly interested in making that distinction before choosing the workshop format. What needs a shared decision, and what needs someone to try it and report back?