When AI use is uneven on a team, I run the weekly meeting in the same shape each time. Fifty minutes, three parts. The point is to avoid rewarding the people using AI more for volume alone, and to avoid treating the people using it less as if they’re just slow.
First fifteen. Each person brings one decision a model couldn’t have made for them this week. Not ‘what are you working on’. A real decision that needed context and judgment from the team. I find the people leaning harder on AI often notice fewer of these until later, because the model doesn’t flag them. The people doing more of the work by hand usually arrive with more of them. Both groups need the question.
Second fifteen. One person walks through one thing they used a model for this week and what changed in the workflow because of it. The speaker rotates. The point is to show where AI is saving time and where it’s creating cleanup or detours, rather than running a demo. The people using it less hear a real account from a teammate instead of a vendor pitch. The people using it more have to explain what changed.
Last twenty. Updates and blockers, the usual business. I put this last on purpose. If it goes first, it expands and eats the meeting. If it goes last, the part that helps the team adapt still happens.
People leave with something they couldn’t have gotten from a model. The people using AI more leave having been asked a question they weren’t asking themselves. The people using it less leave with a clearer read on where the tool helps and where it doesn’t. The meeting keeps the shift in how work gets done visible to humans somewhere and stops one team from quietly turning into two, rather than settling the team’s AI policy.