You sit down to do half an hour on AI. Properly, this time.
Twenty minutes in, you've opened seven tabs you probably won't come back to. You've discovered two tools you didn't know existed. Someone you respect has posted something that makes you feel a half-step behind in a way you can't quite name. You close the laptop.
You're not anxious, exactly. You're just heavier than you were when you sat down. And a thirty-minute task has somehow left you further behind than when you started.
If you've felt some version of this, you're not alone. Helping people navigate uncertainty is my day job, and I'm finding this hard too. What I've slowly come to understand is that this is hard for reasons that have very little to do with who we are as individuals. It's hard because of the conditions we're working inside. And almost nobody is naming those conditions clearly enough for any of us to make sense of what's going on.
AI is different. And for a long time, I couldn't work out why.
What I think now is that AI isn't one kind of difficulty. It's five kinds of hard, all stacked on top of each other, each genuinely hard on its own. And for anyone leading other people through it, there's a sixth weight sitting on top of all the others.
Naming them does something. Not because naming is a substitute for action - it isn't - but because the moment you can see what you're inside, you stop asking "why is this so hard for me?" and start asking a much better question.
1. The ground has stopped settling
Think about the changes you've worked through before. Most of them sat inside an environment that eventually settled. There was a disruption, then a stretch where things calmed and certainty returned, then the next interruption. You got through the hard part, consolidated, operated from a stable base again.
That cycle has gone.
We're no longer moving through episodic disruption. We're living inside sustained uncertainty as the new - and permanent - operating environment. The World Uncertainty Index, which has tracked 143 countries since the mid-1990s, has risen to all-time highs. Above pandemic levels. Above the 2008 financial crisis. Each surge more severe than the last and each recovery shallower.
The floor has moved.
This would be hard to work inside regardless of AI. Most of us are still treating each disruption as an event to get through. But the weather itself has changed.
2. This is a change of an entirely different order
Every change you've led before had edges. A beginning, a middle, some kind of end. You could describe the future state, draw a path to it, bring people with you. The change frameworks we use were built for that - discrete, finite, possible to get your arms around.
I've led real change before. Restructures, mergers, system rollouts, sector reform, crisis response at scale. Most of it was bounded enough that I could see the shape of what I was leading, even when I couldn't see how it would end.
AI isn't that kind of change. It has no edges. There is no defined future state to describe, because the technology is reshaping what's possible faster than any plan can be written. The next iteration arrives before the last has been absorbed.
And this isn't a change that sits neatly inside work. It's already reshaping how we live, how we think, how we decide, how we relate to each other. The change frameworks aren't wrong. We're just operating far outside the conditions they were built for.
3. The learning doesn't finish
By now, you've probably started learning about AI. Maybe in earnest - taking courses, listening to podcasts, trying out different tools. You might feel like you're getting somewhere.
And then - just as you start to approach the point where the learning is supposed to embed and become part of how you operate - new tools appear. Capabilities evolve. What you'd got the hang of has been superseded by something that didn't exist a month ago. You're back at the beginning, in a space that wasn't there when you started climbing.
The learning cycle keeps resetting before it finishes.
This is one of the things that makes AI so draining. Not the doing of the learning - although that's hard too. It's that mastery keeps moving out of reach. Most of us have spent our professional lives inside a model where effort eventually produces mastery. Where learning has a destination, and you can say, at some point: I know this now.
You won't get to say that here. Not any time soon.
AI isn't a mountain to summit. Every time you reach what you thought was the top, you realise you've been climbing the foothills.
4. You don't yet have the words for it
AI is an entirely new domain. New language, new concepts, new ways of thinking about what work even is. You're not extending a skill set you already have - you're learning something genuinely novel, in a space that keeps evolving while you're trying to get your head around it.
You need to know the difference between Gen AI and AGI. What a token is. Why alignment matters. Many of us haven't had to learn at this depth since we were students. The muscle of being a beginner - of not-knowing, of fumbling, of being visibly new at something - has weakened. And we're being asked to use it again, over and over. Publicly.
And it's not just the cognitive load - there's also the emotional load to deal with.
You might be feeling disoriented, excited, confused, curious, overwhelmed. Any of them, or all of them - sometimes in the same afternoon. So you're carrying cognitive and emotional load at the same time, trying to find the words for both.
5. AI is personal in a way previous changes weren't
In the past, a restructure asked you to rethink your career, a new system asked you to learn new processes or a merger asked you to absorb a new culture.
None of them asked you to fundamentally reassess what your value is, now that the boundary between humans and machines is being redrawn.
The first time I watched AI produce in seconds something that would have taken me hours of careful, skilled work, I stopped knowing intellectually that AI was becoming more capable and started experiencing it. Thinking. Writing. Reasoning. Analysing. Things I had understood as the work only humans could do - thinking, writing, reasoning - happening in real time, in front of me, while I sat inside the system being reshaped.
The questions that arrived weren't technical. They were more personal and far more unsettling. If intelligence is now abundant, what value do I offer?
Up until now, our professional identity was something we could shape around our role, our expertise, our competence. Experience - the long arc of things we'd learned to do well over decades - was the scaffold we stood on. Decision-making rested on data, evidence and precedent. What we already knew was the thing that made us useful.
AI is quietly taking that scaffold apart. Experience still matters. But it no longer means what it used to.
Most uncertainty asks what you should do. AI asks who you are.
And the sixth - leading others through all of this as well
Doing all of this for yourself is already an enormous ask. Doing it while leading others through the same terrain is something else again.
Leading people through change has always carried a particular kind of weight. You hold steadiness for others. You make sense of the uncertain bits for them. You absorb the parts of the disruption that don't need to be passed on.
But right now, the territory is regenerating faster than anyone can stay ahead of it. Which means leading others requires you to walk this new terrain whilst you're doing all of it yourself - learning, sense-making, navigating your own questions about identity and value.
You can't lead what you don't understand. And you can't help people through something you haven't begun to navigate yourself.
This isn't a reason to step back from leading. It's a reason to lead differently. With more honesty about what you don't yet know. With more visibility about your own learning. With less performance of certainty, and more willingness to think out loud with the people around you.
Instead of leading from a position of I have worked this out, it's time to lead from the position of I'm inside this with you, and here's how I'm thinking about it.
What changes when you see all of this?
Something shifts when you can see all of this clearly. The quiet, low-grade sense that I should be on top of this by now starts to loosen its grip. Not because anything's been solved - but because you realise that the standards you've been measuring yourself against are no longer relevant.
And you start to recognise that other people are inside these same conditions - and that the difficulty isn't landing the same way for everyone. The colleague who's gone quiet may be stuck inside the learning that won't finish. The team member who seems competent but exhausted may be carrying identity questions in silence. A peer who appears highly confident may be struggling to find language for their internal experience.
The more useful question stops being why is this so hard for me? and becomes: what does it actually take to work and lead well inside this?
If all of this feels hard - you're not imagining it. The real work isn't just learning AI. It's learning AI while learning to live inside sustained uncertainty. Working out what you uniquely bring while the ground keeps moving. Staying human through one of the largest changes any of us will live through. And, for many of us, helping others do the same.
That is the work. And it is learnable.