Groupthink has been hanging around boardrooms for a long time.

We know the basic problem. Put a group of capable people in a room and, under the right conditions, they can become remarkably good at reinforcing one another. Hierarchy plays a part. So does the very human tendency to become a little less certain of our own view when everyone else seems to be nodding.

One of the ways we have tried to protect against this is independent preparation. Read the papers. Think the issue through beforehand. Form a view before you hear everyone else’s.

I use AI to think with before important conversations. Quite a lot, actually. It helps me test an argument, find something I have missed, take another position seriously, or work out why I am still unconvinced by something that looks perfectly reasonable on paper.

And this has left me wondering whether one of our protections against groupthink is becoming rather more complicated.

Five years ago, asking people to prepare independently was fairly straightforward. Now, a leadership team can arrive at a meeting having each spent time, separately, thinking about the same issue with AI.

They may use different systems and ask different questions. They still bring their own experience, judgement and knowledge. But there may also be more common influence in the room than we can see.

AI can affect what feels important, which risks stand out, how the problem gets framed in the first place and which possibilities get considered. We don’t yet know enough to say what this is doing to the diversity of thinking in executive teams. But it does make the word independent a little less straightforward.

I’ve started thinking of this risk as Modelthink.

Not because AI is simply producing groupthink by another name or another route - that would be too easy. The more interesting possibility is that apparently independent thinking may have more common shaping in it than we realise.

And then there is preparation itself.

AI makes it possible to prepare extremely well. We can anticipate objections, test weak points, rehearse difficult questions and keep working at an argument until we are very clear about what we think.

Perhaps too clear.

Because there is a point at which being well prepared can become arriving with very little left open.

I’ve been thinking of that risk, as Closedthink. It doesn’t look closed-minded. Could be quite the opposite in fact. The person may have tested the position from every angle, considered the counterarguments and worked the reasoning through carefully.

They are ready.

The question is whether their thinking is still available to be changed.

One of my academic heroes, Mary Parker Follett was interested in exactly this quality of human interaction more than a century ago. Her idea of circular response was that we affect one another through the process of meeting. We don’t simply bring fixed positions together and compare them; what happens between us changes what each of us sees.

That, to me, is part of what a good meeting is for.

I want to bring a view into the room, and I want you to bring yours. But I also want something to become possible between us that neither of us could have arrived at alone.

Your question may change what I notice. My experience may alter how you understand the problem. Something I thought was settled may turn out not to be. Something neither of us could see alone may only become visible once we are looking at it together.

And this is where the old concern about groupthink becomes more interesting.

We still need to notice when the group is influencing us too much. Now we may also need to notice whether our supposedly independent preparation has become more alike than we think, and whether we have prepared ourselves so thoroughly that other people can no longer influence us enough.

I still think we should come prepared.

I’m just less certain that prepared means the same as finished.