The Great Flattening
AI can absolutely help us manage more people. But does that necessarily mean we can meaningfully lead more people?
By now, we've all heard about the great flattening happening across tech. Fewer layers, wider teams, faster decisions. The promise is that organizations will become more agile, with less bureaucracy standing between people and the decisions they need to make.
I'm a believer in fewer layers.
I've seen firsthand how being buried under too many layers can leave people feeling disconnected from decisions, less empowered to act, and further away from the larger strategy.
So the argument for flattening makes sense to me.
But I've been thinking about something else. How are companies deciding what to flatten, and how far to go?
Much of the conversation seems to come down to span of control — how many people report to a manager. If a manager has three direct reports, could they manage six? Or ten? And with AI taking on more operational and coordination work, could that number go even higher?
Perhaps.
But I think we're combining two different questions into one.
How many people can a manager manage? And how many people can a leader meaningfully lead?
Managing involves coordination, tracking progress, removing blockers, and keeping work moving. AI is already making some of this easier, and I expect it will continue to improve.
Leadership is different.
It's understanding what motivates someone, recognizing potential they may not see in themselves, helping them navigate ambiguity, and creating the conditions for them to grow. It's knowing when to step in and when to step away.
And that takes attention. Not just time on a calendar, but the ability to really understand the people you're leading.
I don't think there's a universal number that defines the right span of control. It depends on the nature of the work, the maturity of the team, the level of autonomy, and what we expect from managers in the first place.
Which brings me to the question I've been sitting with.
If AI changes how much work a manager can handle, shouldn't we also rethink what we expect a manager to do?
Otherwise, we risk using AI to make existing management structures more efficient without fundamentally improving leadership.
Maybe the opportunity isn't simply to give managers more people.
Maybe it's to use AI to create more capacity for the parts of leadership that have always mattered most.
And that, to me, is a more interesting way to think about flattening.