The “Great Flattening” Is Here. But Are We Removing Managers Faster Than We Develop Leaders?

business2026-08-28 · 1 min read · 1.3k reads

AI is helping companies simplify structures, widen spans of control and rethink middle management. I believe much of that change is necessary. But after more than 20 years working with people and organizations, I think there is another question we need to ask: what happens to human development when

The Great Flattening is no longer just a prediction

Something important is happening inside organizations.

As AI takes over more reporting, coordination, information processing and administrative work, companies are beginning to question how many management layers they actually need.

And perhaps they should.

The conversation has even acquired a name: the “Great Flattening.”

Across the technology sector, companies are experimenting with wider spans of control, fewer management layers and new “player-coach” models in which managers are expected to contribute directly to the work while leading increasingly large teams.

A Business Insider report published this week describes managers overseeing significantly larger teams while simultaneously being expected to remain hands-on. The consequence is not difficult to imagine: when immediate delivery takes priority, coaching and long-term career development can be among the first things to disappear.

Another recent analysis points to an equally important problem. Research cited by HR Executive found that only 6% of leaders surveyed believe their middle managers are prepared to lead change across most of their organizations. At the same time, managers are being asked to handle wider spans of control, AI-enabled workflows and increasingly complex expectations.

So perhaps the question is no longer whether organizational structures will change.

They already are.

The more interesting question is:

What are we removing together with those layers?

The “Great Flattening” Is Here. But Are We Removing Managers Faster Than We Develop Leaders?

“Omul sfințește locul.”

In Romania, we have an expression that is difficult to translate perfectly:

“Omul sfințește locul.”

Literally, it means something like “the person makes the place.”

But its meaning goes deeper.

To me, it means that a place, a team or an organization ultimately becomes what it is because of the people inside it.

Because of how they show up.

How they treat one another.

What they build together.

What they tolerate.

What they encourage.

And what they make possible for others.

I think about this expression increasingly often when we talk about AI and the future of organizations.

Because an organization is not simply a system for distributing tasks.

It is a human environment.

People need information, certainly.

But they also need understanding.

They need feedback.

They need someone who notices when they are struggling.

They need difficult conversations and honest conversations.

They need to watch someone more experienced navigate a situation they have never faced before.

They need encouragement at one moment and challenge at another.

They need opportunities to make mistakes, learn from them and try again.

Sometimes they simply need another human being to say:

“I've been here before. Let me help you think through it.”

AI can support many of these moments.

But supporting human interaction and replacing it are not the same thing.

What happens when efficiency removes the spaces where people used to learn?

This is the part of the Great Flattening conversation that interests me most.

A manager is not valuable simply because there is a box with “Manager” written on an organizational chart.

And we should be honest: not every management layer creates value.

Some create complexity.

Some slow decisions.

Some exist because organizations have grown without ever redesigning themselves.

Flattening those structures can make perfect sense.

But good managers do something much harder to capture on an organizational chart.

They create context.

They translate complexity.

They help people understand not only what decision was made, but why.

They recognize potential before it appears in a performance dashboard.

They handle conflict.

They give feedback.

They share experience.

They create accountability.

They notice when someone is ready for more.

And, perhaps most importantly:

They create the conditions in which other people can grow.

Recent research is beginning to identify exactly this problem from another direction. As AI absorbs more routine early-career work, organizations risk removing some of the tasks through which younger employees historically developed judgment and expertise. McKinsey recently argued that the informal mechanisms that once turned novices into experts can no longer be taken for granted and that organizations will need to redesign learning, coaching and early-career development more deliberately.

That is not a small organizational-design issue.

It is a future leadership pipeline issue.

The real risk isn't fewer managers

I don't believe organizations need to preserve management roles simply to preserve management roles.

Some will disappear.

Some should.

Others will change dramatically.

The real risk is different:

We may remove opportunities to learn leadership faster than we create new ones.

Think about how many leaders developed historically.

Someone was given responsibility for a small team.

Then a larger one.

They watched a more experienced leader.

They made mistakes.

They handled their first difficult performance conversation.

Their first serious conflict.

Their first organizational change.

Their first budget.

Their first crisis.

Their first decision where there was no obviously correct answer, but people were still waiting for them to decide.

Over time, those experiences accumulated.

That is how judgment develops.

Not from information alone.

From experience + reflection + feedback + other people.

AI is already changing some of these traditional pathways. The World Economic Forum recently highlighted the concern that automation is removing some of the work that previously acted as a testing ground for employees progressing toward leadership, while emphasizing the need for organizations to deliberately redesign work and reskilling around human-centered capabilities.

Which leads to a question I believe every leadership team should be asking:

If we remove layers from the organization, where will the next generation learn to lead?

AI should give us more space to be human, not less

This is where I see enormous opportunity.

If AI can eliminate hours of reporting, administration, coordination and repetitive analysis, that should be good news for managers.

But then I would ask:

What are we doing with the time we get back?

Are we simply increasing spans of control?

Reducing headcount?

Adding more tasks?

Expecting one manager to supervise 20 people instead of 10?

Or are we using some of that capacity for the things organizations have been telling us for years they don't have enough time for?

Coaching.

Mentoring.

Listening.

Thinking.

Developing people.

Building stronger teams.

Having difficult conversations properly.

Helping someone prepare for their next role.

Creating space for experimentation.

Helping people navigate change.

Because perhaps one of the greatest opportunities AI gives organizations isn't simply to remove people from work.

Perhaps it is to remove enough low-value work that people finally have more time for the work that only people can do.

And that, to me, is a much more interesting future.

We may need to redesign leadership development itself

If organizations become flatter, leadership development cannot continue to depend primarily on promotion into management.

We will need to create experience deliberately.

That might mean giving people earlier opportunities for:

project leadership;

cross-functional responsibility;

customer exposure;

mentoring and reverse mentoring;

shadowing senior leaders;

decision-making responsibility;

ownership of real business outcomes;

and opportunities to lead before receiving the formal title of “manager.”

The manager of the future may also need a very different capability set.

AI fluency will matter.

But so will judgment.

Empathy.

Context.

Communication.

Trust.

The ability to navigate ambiguity.

The ability to lead people through uncertainty.

And the ability to know when technology should make the decision easier, and when a human being still needs to make the decision.

Recent McKinsey work on AI transformation points in the same direction: AI adoption is moving faster than many leadership models can absorb, with leadership readiness and change management emerging as constraints on successful adoption.

Technology can scale very quickly.

Human capability takes longer.

That difference matters.

After more than 20 years, this is what I believe

I have spent more than two decades working with people, leaders and organizations.

I have seen strategies change.

Structures change.

Leadership teams change.

Processes change.

Technology change.

I've seen organizations that looked excellent on paper struggle enormously in practice.

And I've seen imperfect organizations achieve extraordinary things because the people inside them created trust, accountability, energy and a willingness to build something together.

That experience has left me with a very simple belief:

People need people.

We need experience that can be shared, not only information that can be accessed.

We need empathy.

We need someone to challenge us.

Someone to believe in us.

Someone to tell us when we are wrong.

Someone to help us see what we cannot yet see ourselves.

We need conversations that don't fit neatly into a workflow.

We need environments in which people can learn, fail, recover, develop and eventually help someone else do the same.

AI will become extraordinarily good at many things.

And I believe we should use it.

I want organizations to automate unnecessary work.

I want leaders to have better information.

I want people to spend less time on repetitive tasks.

I want organizational structures to become simpler where complexity adds no value.

But as we redesign organizations around what technology can do, we should be equally deliberate about protecting and strengthening what people do for one another.

Fewer managers may make sense. Fewer leaders do not.

Maybe the organization of the future really does need fewer managers.

I am not convinced it needs fewer leaders.

And I certainly don't believe it needs less humanity.

So before removing the next organizational layer, I would ask leadership teams one more question:

What work are we eliminating, and what human experience, learning and development might we accidentally eliminate with it?

Because becoming more efficient is relatively easy to measure.

Becoming a better organization is something else entirely.

What do you think?

As organizations become flatter and AI takes over more coordination and administrative work, are we creating enough new ways for people to learn, develop and become leaders?

I'd be interested in hearing different perspectives, including those who disagree.

I write about Human Capital, leadership, organizational transformation, AI and the future of work. If these are conversations you care about too, you can subscribe to my AVALW page to receive my next articles.

Sources & further reading