Are You Up for This?

Why AI Leadership Is Really a Test of Capacity 

by Mike Herzog, President, the telos institute

This essay is the second in a series exploring how AI is destabilizing the ways organizations understand themselves, and what kind of leadership may be required when the old frames no longer hold.

2 a.m.

As you scroll through your LinkedIn notifications, it feels like a normal day.

A few work anniversaries warrant the obligatory “Congratulations.” A distant colleague announces a promotion. You see several posts that feel, at least to you, like ego pieces dressed up as thought leadership. Even if you wanted to read them, who has time?

Then you see it.

It registers in your gut before you have fully taken in the words.

Your biggest competitor has just announced a new AI-based initiative. Damn, this looks big. Worse, it’s impressive. Your intuition tells you this may be a game-changer.

How long have they been working on this? Surely, a long time. How long would it take your team to bring something like this to market? Are you even capable of something this innovative?

Part of you wants to scroll past the post and push it into the rear-view mirror of your mind. Instead, you hit the like button and type a brief “well done,” even though it kills you to do it.

For the rest of the day, it is easy enough not to dwell on it. You throw yourself into the immediate work. Internal meetings. Financial reports. Client reach-outs. The small decisions that somehow consume the whole calendar. If you are lucky, you might even get some real work done.

Before you know it, the day is over.

Then it is 2 a.m., and you are wide awake.

At first, the thoughts feel random. Disconnected. A loose swirl of anxiety, strategy, resentment, curiosity, and dread.

How long have they been working on this? What do they know that you don’t? Is this real, or just a well-packaged announcement? Are you already behind? How far behind? Is your team moving fast enough? Are you? What are you not seeing?

Slowly, the pattern reveals itself.

You’ve been wrestling with this for eighteen months, but somehow it still feels new. Who would have thought two letters could carry this much uncertainty? Could create this much pressure?

You have seen disruptions before. You have faced downturns, restructurings, competitive threats, technology shifts, and more urgent strategy decks than you can count. But this feels different.

Yes, there is the technology itself. But you’ve dealt with new technology before. This feels different. Personal.

You cannot simply delegate this to the smart people on your team and trust that they will figure it out. AI reaches into every part of the business. It touches strategy, operations, people, risk, knowledge, judgment, client value, and the way work itself gets done. It reaches all the way down into your own day-to-day habits.

And it is not just the technology. It is what the technology exposes.

It’s easy to say that efficiency will become innovation, and innovation will become growth. Maybe it will. But what if AI does not simply make your organization faster? What if it puts pressure on the places where you are already weakest?

Then there are the people.

You hear the whispers. People are watching the hiring slowdowns and layoffs. They are reading the same headlines you are reading. They are trying to understand what this means for their roles, their value, their future.

They want comfort. They want certainty.

Here’s the problem.  You are uncomfortable.  And you are far from certain.

That is why you are awake at 2 a.m.  What exactly scares you most?

Move too slowly, and your competitors may beat you to market. Move too quickly, and you may become reckless. AI is overhyped. AI is the most profound change we have seen in a generation. It is both. That is part of the problem.

For everything it is, there is one thing it is not.

Clear.

There are no solid footholds. No safe paths forward. No clean way to separate threat from opportunity, speed from wisdom, experimentation from irresponsibility.

All of that is true. But none of it is the real reason you are lying awake.

The thing that scares you is you.

Are YOU up for this?

You have always been up for a challenge, so why does this one feel different?

Because this challenge is not only asking you to decide. It is asking you to become.

Your intuition is telling you that the complexity in front of you requires more than a better strategy, a sharper operating model, or a faster adoption plan. It requires more from you personally. More capacity for ambiguity. More tolerance for paradox. More ability to act without pretending you are certain. 

You are in liminal space.

A threshold between two realities.

The rules from the old world no longer fully apply. The rules of the new one have not yet formed. And there is no going back.

This is, of course, about the business. The business is in that threshold too. So are your competitors, your clients, your employees, and the broader society around you.

But at 2 a.m., that is not the part you can avoid.

This is about you.

It is about the leader you must become in order to lead when the old frame no longer holds.

When Solving Fails

You have spent your entire career solving hard problems. 

Figure out what is not working. Devise a solution. Get the right people focused on it. Move the work forward. Your experience and the right team around you have usually been enough. 

With enough data and enough discipline, you could get to the right answer.

So what is different this time?

AI is not just another hard problem to solve. Yes, there are policies to write, tools to evaluate, risks to manage, and people to train. Those problems are real. But beneath all of them, something larger is happening. 

The paradigm is shifting while you are still inside it.

AI reaches almost every assumption your organization has made about how work gets done: what counts as expertise, where judgment lives, how value is created, what stays human, which roles still make sense. 

The disruption is obvious. The end-state is not. 

Nobody knows exactly where this is going. Not the vendors. Not the futurists. Not your competitors. Not you.

And when the paradigm is changing, the past no longer serves as a reliable guide. Data and analysis still help. Strategy and expertise still matter. But none of them can resolve the situation, because the situation is not only complicated. It is still forming.

That is why solving begins to fail. Not because there is nothing to solve, but because solving is too small for the moment. It drives progress inside an old frame while the frame itself is coming apart.

So the leadership question changes. How do you lead when the work is not to solve the problem, but to understand what kind of problem you are actually in?

Most leaders do one of two things.

Some wait. They watch the market, track the headlines, let others experiment first, then move quickly once the dust settles. 

But your intuition tells you this one is different. AI is moving too fast, and the advantage your competitors gain may not be a product advantage. It may be a capability advantage. Waiting carries its own risk.

The other option is to act. But not all action is the same. There is action designed to stabilize, and action designed to discover.

When leaders find themselves in liminal space, most reach for solid ground. Move decisively. Solve what can be solved. 

There’s nothing wrong with this. Much of it is necessary. The captain in the storm is an appealing image: calm voice, firm hand, clear direction. Employees want it and boards reward it.

But what if the work is not steering the ship? What if the work is seeing that the waters have changed, the instruments are unreliable, and the map was drawn for a different world?

That’s the hidden danger of stabilizing action. It looks like leadership while reinforcing the existing paradigm. It creates motion and visible progress while avoiding the deeper question AI is placing in front of you. 

It turns AI into a familiar category.

A technology project.
A productivity initiative.
A cost-reduction opportunity.
A threat to be managed.

Each may be partly true. None may be large enough.

That is where discovery becomes essential. 

Discovery is not reckless experimentation. It is taking action in order to learn what analysis alone cannot tell you. The leader still decides, but the purpose shifts. The goal is not to restore certainty as fast as possible. It is to learn what the uncertainty is revealing.

Rather than asking only, "How do we make this process faster?" they ask, "Why does this process exist at all?" Rather than only, "Where can we automate?" they ask, "Where does human judgment matter more than we realized?" Rather than only, "How do we keep up?" they ask, "What kind of organization are we becoming as we try?"

These leaders act, but they do not confuse action with certainty. They know some paths will dead-end, and they build that in: financially, operationally, and emotionally. They do not promise they know where this ends. They promise to learn faster and more honestly than they could by standing still.

That kind of leadership sounds reasonable. So why is it so rare?

Because discovery sounds simple from the outside. Act to learn. Stay curious. Don't over-control a future you don't yet understand. But most leaders cannot sustain it because discovery asks them to tolerate what they have not yet developed the capacity to tolerate. 

Ambiguity without collapse.
Motion without certainty.
Progress without the comfort of a fixed destination.

That is why AI is not only testing strategy. 

It is testing something more fundamental: the leader's capacity to meet complexity without reducing it too quickly.

The Wrong Problem

Much of the conversation about AI rests on a single assumption: that AI is primarily a technology problem.

A complex, fast-moving, consequential one. But a technology problem all the same.

That assumption is not exactly wrong. It is just too small.

AI is the forcing function pushing organizations into liminal space. The deeper challenge sits at the level of leadership, and it is developmental.

If AI were only a complex technical challenge, the response would be mostly informational. More knowledge. Better tools. Sharper implementation. Stronger governance. Smarter strategy. The job of leaders would be to climb the learning curve faster than the competition.

A challenge, but not the kind that keeps someone like you awake at 2 a.m.

What keeps you awake is the intuition that this is more than technical.

AI is not only asking what you need to learn.  It is asking who you need to become.

Leaders do not merely need to choose the right AI platform faster than their competitors. They need to stand in uncertainty long enough to choose well.

They do not need to know the most about AI. The half-life of that knowledge is measured in months. They need to act without waiting for an answer that may not arrive.

They do not need a confident roadmap to the destination. They need to move with real conviction toward a future they cannot yet see, and bring people with them without pretending they can see it.

And they do not need only to steady the ship. They need to stay steady themselves when the instruments fail and the map no longer matches the water.

Notice what these require.

Not just skill. Not just knowledge. Not just confidence.

They require the ability to stay present with complexity, contradiction, uncertainty, and emotional pressure without reducing them too quickly into a simpler story.

They require a leader who can remain open when the organization wants closure, act without using action to escape discomfort, and offer steadiness without pretending the situation is clearer than it is.

That is capacity.

The capacity to hold more complexity without reducing it too quickly. The capacity to remain open when the organization is asking you to close things out. The capacity to act without using action as a way to escape discomfort. 

That is not what most leadership development has spent the last thirty years building.

We have trained leaders to know more, decide faster, communicate better, and project certainty. Those things still matter. But we have spent far less time developing their ability to stay present in the face of what they cannot yet resolve.

That is the gap AI is exposing.

Not only a gap in what leaders know. A gap in who they have been developed to be.

And here is the difficult part. Every one of those capacities is simple to describe and hard to embody.  Knowing that you should stay open in uncertainty does almost nothing to help you do it at 2 a.m.

Understanding the gap is not the same as closing it.

The Opening

The capacity you need is not something you can simply decide to have.

You cannot read your way into it, acquire it from a workshop, or resolve at the start of the quarter to become more comfortable with ambiguity.

And it is not a matter of getting more information. You already have more than you can use.

The real constraint is how much you can hold in view at once. How much complexity, contradiction, risk, emotion, and uncertainty you can keep open before the discomfort forces you to narrow it back down to something more manageable.

That is the aperture through which you take in the world.  The question is not only how much information reaches you.  It is how wide the opening is.

That aperture can widen. But not quickly, and not by trying harder.

It widens the way any deep capacity grows: over time, under real pressure, usually with help, and almost never alone. This is the slow work the AI conversation keeps skipping past because it cannot be bought in a license, delegated to a task force, or installed in a quarter.

It is the work of becoming a larger leader than the one who walked in.

And here is the part worth sitting with.

That widening does not happen after you leave the threshold.  It happens because you are in it.

The uncertainty keeping you awake is not only the problem to be managed. It is also the condition in which this kind of growth becomes possible.

Liminal space is not only the disruption.  It is the opening.

The Work at the Threshold

This is why the real AI conversation cannot stop at strategy.

Strategy matters. Governance matters. Tools, use cases, and risk management all matter. But none of them answers the deeper question AI is placing in front of leaders.

Not: what will we do with this technology?  But: who must we become to meet what this technology is revealing?

Organizations still have to move. Decisions still have to be made. People still need direction, protection, and meaning.

The leader's job is not to manufacture certainty before acting. It is to move in a way that deepens understanding rather than shutting it down, and to offer steadiness without pretending to be certain.

So come back to the question that found you at 2 a.m.  Are you up for this?  If the question means “do you already have what this requires,” then no.  You don’t. And neither does anyone else.

AI did not arrive with a manual, and the people who sound most certain about where it leads are guessing too.

But notice that you are still here.  Still awake. Still holding the question open instead of collapsing it into the nearest answer just to make the discomfort stop.

That is not a small thing.  In a moment crowded with false certainty, the willingness to stay in the question is a rarer capacity.

The fear you felt was not weakness. It was an accurate reading of the situation: what is in front of you may require a larger version of you than the one that got you here.

That is the threshold.

You cannot see the other side, and there is no going back.

The only question is whether you will cross it by default, shaped mostly by pressure and luck, or with enough intention to let the uncertainty change you.

There is no leading around it.

Only through.