This page was translated from Dutch with AI.
Future of work
What is cognitive agility? Five human capacities that AI cannot automate, from flexible thinking to intuition. By Steff VanHaverbeke.
The future of work is not a race against the machine. It is a journey back to what makes us human.
When AI takes over the routine work (the emails, the first drafts, the number crunching, the planning), something fundamental shifts. What remains becomes more valuable. Attention becomes scarce. Judgement becomes precious. Being able to live with ambiguity, sensing the mood in a room, or making a decision that has no clear algorithm behind it: these become the new hard skills.
For decades we have optimised everything for efficiency. Now efficiency is becoming cheap. The question is no longer "how do I do more?" The question is "what is actually worth doing?"
Cognitive agility is the uniquely human ability to adjust your thinking and your emotional and social approach, so you can find your way in messy, unpredictable situations. It is not a single skill, it is a meta-capacity. The ability to break patterns, look at problems differently and make decisions where data and a deep human awareness come together.
I have identified five capacities that together form the core of cognitive agility. Together they form the CAF, the Cognitive Agility Framework. These are not abstract concepts. You can see them, you can train them, and they are becoming increasingly valuable now that AI is taking over the predictable work.
The pattern breaker
Metacognition, systems thinking, reframing
The ability to switch between analytical, creative and systems thinking, and knowing which mode suits the moment. Metacognition, or thinking about your own thinking, is the key. AI works with patterns from known data. Humans can pause, take a step back and ask a completely different question.
The trust builder
Psychological safety, silent data, presence
Not just understanding emotions, but also building the trust and psychological safety that make honest work possible. Picking up the "silent data" in a room. Knowing when someone's silence means agreement and when that person has already checked out. AI has no body, no lived experience and no stake in the relationship.
The co-creator
Shared meaning, alignment, co-creation
Going beyond coordination, towards true co-creation. Shared files are not yet shared meaning. Real collaboration requires presence, alignment and the willingness to let the conversation change you. AI can facilitate. Humans can collaborate.
The pattern sensor
Embodied knowledge, gut feeling, experience
The embodied knowledge that comes from decades of experience: the gut feeling that something is wrong before you can say why. Intuition is not the opposite of data. It is what happens when experience comes together faster than your conscious reasoning can follow. AI finds patterns. Humans feel them.
The reimaginer
Conviction, courage, emergence
Not just coming up with ideas, but creating the conditions in which new things can emerge, even in systems that resist change. Innovation requires human conviction, cultural context and the courage to pursue something for which there is no proof yet. AI can remix. Humans can reimagine.
This is the part of the AI conversation that gets too little attention.
When you outsource cognitive work to AI (the research, the writing, the analysis, the decision support), something subtle starts to happen. The muscles you do not use begin to waste away. Not spectacularly, not overnight. But steadily.
This is what I call The Drift in Finally, Superpowers!. It is not a technology problem. It is a problem of human attention. The solution is not to stop using AI. The solution is to consciously choose which thinking work you keep for yourself: the work that sharpens your judgement, not just the work that feels productive.
In the compute economy, machine cycles are abundant. Computing power becomes cheaper every year. But there is one form of computing power that remains scarce: human attention, human care, human judgement.
| Machine compute: abundant, becomes cheaper every year and scales endlessly | Human Compute: scarce, you cannot manufacture attention, care and judgement |
|---|---|
| Processing data at scale | A doctor looking a patient in the eyes |
| Recognising patterns in dashboards | A manager sensing burnout before the numbers show it |
| Optimising for short-term numbers | A leader choosing long-term trust over quick wins |
These are not soft skills. They are the scarce resources in a world of abundance. And they are what organisations will pay a premium for, now that AI handles everything else.
If you lead a team today, you face pressure on three fronts at once.
If one person with AI can do what an entire team used to do, everyone's expectations about output change. But the working day does not get shorter. The workload simply gets bigger. Leaders need to be honest about this, instead of pretending it is pure productivity gain.
Work becomes harder to see. If someone in your team builds a workflow that automates half of their role, does that person become more valuable or less visible? The old metrics (logged hours, sent emails, attended meetings) no longer measure what matters.
For many leaders, expertise was authority. "I know how to do this" was the source of credibility. But if AI knows it too, the value of a leader shifts from knowing to making meaning, from answers to better questions.
I explore these ideas thoroughly in Being Replaced: The Five Human Skills AI Can't Automate. Not just another AI hype book, but a sober look at what makes humans irreplaceable, based on neuroscience and tested in real organisations.
Hi! I'm this site's assistant. Ask me anything about the work, the books or how to work together.
Answers are AI-generated from this site and may contain mistakes.