Degree 6 · Unit 6.3
Leading hybrid teams
Leadership in a hybrid team does not change in its principles. What changes is where the friction sits, and there are four points of friction that appear in every team I have ever seen.
1What are you appraising now?
When producing a draft took two days, the act of producing one was itself a measure. Today a beginner produces ten drafts before lunch. So the measure has moved to the things the tool does not produce: the quality of judgement, meaning what they rejected and why; the rigour of their verification; the framing of the problem; and their willingness to take responsibility. Managers who go on appraising by quantity end up rewarding the weakest people on their team.
2The gap inside the team
You will find three groups in front of you. There are the eager, who use everything without much caution. There are the resisters, who refuse all of it. And there are the silent ones in between, who use it and do not say so — the largest group of the three, and the most dangerous wherever there is no policy. You do not treat this with an enthusiasm campaign, and you do not treat it with a blanket ban. You treat it with three things: clarity about what is permitted, safety in declaring what you use, and room for the resister to learn without being embarrassed.
Where does AI alone suffice, and where is a person indispensable? This is the matrix you redistribute your team's work with.
The practical rule: the higher the risk or the emotional weight, the larger the human share — never the other way round, however tempting the efficiency.
3Where does the next expert come from?
This is the gravest matter in the unit, and the one least discussed anywhere. Expertise gets built out of exactly those simple, repetitive tasks that we now hand straight to the tool. And if we give a beginner a finished result and ask them to review it, when exactly are they supposed to build the instinct that would make their review worth anything?
The answer I would recommend is a deliberate formation path. Have the beginner produce the work themselves first, then compare what they produced with the tool's output, and then explain the difference between the two. It is slower in the first month and faster within the first year. Skip it, and in five years you will have a team that reviews work without really knowing what it is reviewing.
AI is a team sport — no one person masters every link in it. These are the roles the next expert is formed among.
Optional roles are added as you mature: a user-experience designer, a domain expert, a researcher. The most dangerous pattern is leaning on a single star who holds every role — a bottleneck, knowledge locked in one head, and collapse when they leave.
4Lead by what you do
Nothing embeds a policy like a team watching you keep to it when keeping to it costs you something. Say openly where you used the tool, where you refused to, and where it got something wrong on your behalf. A leader who hides their own use produces a team that hides its use — and that is the most dangerous thing that can happen at this stage.
Do this
1 — In your field. Rewrite the appraisal criteria for one role in your team so that they reward judgement and verification rather than quantity.
2 — In practice. Design a formation path for one beginner: what do they produce themselves in the first three months before they call on the tool?
3 — As a leader. At your next meeting mention one place where the tool got something wrong on you and how you caught it.
Where to after this unit? You have led your team. The next unit widens the view to a market and an economy shifting under our feet.