The Thinking Leader: Metacognition in the Age of AI
The Thinking Leader. Credit: Rich Baker
Edition Three.
In the first two editions, we explored how generative AI is reshaping internal communication -from writing with impact, to preserving editorial quality in an age of automation.
This time, we’re shifting focus. Not to the tools. But to you.
Because while AI can support the work, only humans lead it. And the leadership advantage isn’t about using AI faster. It’s about thinking better.
This piece is for anyone in a position of influence - communicators, HR leaders, managers, senior execs - who finds themselves wondering not just what to do next, but how they’re deciding what to do next.
It’s for those who are (at least sometimes) quietly resisting the pressure to perform, and instead choosing to pause, reflect, and lead with intention. And it’s written with a simple premise in mind:
In a world full of action, the rare skill is reflection. What follows is not a productivity hack, or a wellness sermon. It’s an exploration of metacognition - the ability to observe your own thinking - and why it might just be the defining leadership capability of the AI era.
What is metacognition?
Let’s start with the term, because it’s one of those words that manages to sound both academic and slightly made up at the same time.
Metacognition means “thinking about thinking.”
It’s your ability to observe your own mental processes - to notice your assumptions, question your habits, and understand how you’re reaching your conclusions, not just what those conclusions are.
Psychologist John Flavell coined the term in the 1970s, defining it as our awareness and control of our cognitive processes. It’s what lets you pause, mid-email, and ask: “Am I reacting here? Or responding?”
In leadership terms, it’s the internal shift from “What should I do next?” to “What’s shaping how I’m deciding what to do next?”
Rest assured it’s not abstract. It’s not theoretical. So you can carry on reading. In my opinion, it’s the single most important internal muscle a leader can develop in the age of AI.
Why now?
Let’s be honest: AI is getting good. Very good. Very Very Good. Very Very Very - you get the drift.
It can generate content, automate analysis, draft responses, summarise documents and restructure workflows. It can code, create apps, draw pictures, generate videos, and it once cited three impressive-sounding initiatives as examples of my past work. I’d never heard of any of them.
Why?
All the stuff that used to make us look impressive in meetings? It can do most of it now. Faster. Cheaper. (Though it does use a worrying amount of water.) And it never takes a lunch break. Which means our real value isn’t in doing the thing; it’s in knowing what matters, when to do it, and why. AI can give you outputs. But only you can notice when you’re making decisions from bias, assumption, or fear.
“When AI handles the execution, the remaining work is about discernment.” MIT Sloan, 2023
Or to borrow from Daniel Kahneman:
AI thrives on System 1: fast, instinctive, automatic. But leadership belongs to System 2: slow, deliberate, effortful. Metacognition is how we choose which system we’re in, and when to switch.
Mindfulness ≠ Metacognition
You’d be forgiven for thinking mindfulness is the same as metacognition. However, they’re not the same. Mindfulness is about presence: being aware of what’s happening now, without judgement. Metacognition is about perspective, stepping back from what you’re thinking and noticing the patterns.
Mindfulness might help you notice you’re feeling annoyed in a meeting. Metacognition helps you realise you’re projecting last week’s frustration onto this conversation, and perhaps you’re about to overreact. One’s about noticing. The other’s about adjusting. They’re both useful. But metacognition is what gets you out of your own way.
The cost of not thinking about thinking
Without metacognition, we don’t lead - we react. We become decision machines with too much input and not enough clarity.
We’ve made pace a proxy for progress. And in a world where speed is everything, we mistake it for effectiveness. We roll out the tool, push out the update, send out the message all without asking: “What’s really going on here?”
“When we cannot see our own thinking, we are doomed to repeat it.” — Jennifer Garvey Berger
And if that is starting to feel a bit too theoretical, try this: if an AI replicated all your decisions from the past six months would that worry you?
That’s not a tech question. It’s a thinking one.
What metacognition looks like in leadership
This isn’t about retreats and journaling. It’s about real, observable pauses. Moments where you catch yourself before default kicks in. It might sound like:
“What’s the story I’m telling myself about this person?”
“What am I assuming is true, and how do I know?”
“Is this urgency real, or just loud?”
“Who benefits if I believe this version of events?”
It’s not about having the perfect answer. It’s about noticing the question before you plough ahead. Leadership isn’t about managing inputs. It’s about mastering awareness. It’s about holding space for complexity, ambiguity, and doubt - without flinching.
Metacognition helps you do that.
AI sharpens the need - not the edges
The more AI shows up in your workplace - writing drafts, suggesting routes, offering analysis - the more tempting it becomes to stop thinking critically.
That’s the trap.
You start rubber-stamping AI-generated messages. You let automation dictate priorities. You become the human in the loop… without realising the loop’s doing all the leading.
But leadership isn’t being in the loop. It’s being above the loop.
Noticing the system. Naming the tension. Asking the better question. AI can’t replicate the slow, layered discernment only a human mind, beautiful in its uncertainty, can offer.
Metacognition as emotional intelligence, version 2.0
A lot of this is emotional, not just intellectual. Metacognition gives you:
Emotional regulation (you don’t lash out mid-meeting)
Cognitive flexibility (you update your views when the facts change)
Self-awareness (you see the patterns you fall into)
It’s what helps you hear criticism without collapse. To respond with curiosity instead of defensiveness. To spot when your brain’s chasing certainty because ambiguity feels a bit scary today.
AI won’t help you with that. But you can. And the more AI does the operational stuff, the more this is the work.
So what do we do with this?
We don’t need another 5-step framework. We need to practise noticing. Here’s a suggested starting point:
In your next meeting, watch your thinking.
Ask yourself why you’re drawn to one idea more than another.
Pause before replying. Not to delay - but to consider what’s shaping your response.
If something winds you up, ask: “Is this about them… or about me?”
It’s not about being perfectly rational. It’s about knowing when you’re not.
“Learning how to think really means learning how to exercise some control over how and what you think.” — David Foster Wallace
And if you’re designing leadership development programmes, this matters even more. Too many still centre around classroom exercises - models, frameworks, case studies.
But the kind of leadership we need now? That’s a life exercise. Metacognition isn’t a module. It’s a muscle. One that needs practice in real meetings, real emails, real moments of “Oh, hang on - why am I reacting like this?”
Great coaching supports it. But the work is yours. This isn’t about having someone else unlock your thinking. It’s about learning to do it yourself.
Final thought
AI has changed what leadership looks like. But it hasn’t changed what leadership is. And if that unsettles you just a little - good. Because leadership isn’t static. It’s active. And so is how we think.
And right now, the world doesn’t need louder leaders. It needs thinking ones.
The ones who can hold a question. See the pattern. Step back before stepping in. Because the tools are levelling the playing field.
And the only unfair advantage left will be how well you think.
Thanks for reading.
