The question is
what to learn.
Or perhaps:
what remains scarce?
Most people ask, “What skills should I learn in the age of AI?” I reopened that question seven times. Each time, a different way of seeing the problem emerged.
“What skills should I learn in the age of AI?”
It is the most common question. The answers seem obvious:
AI fluency, data literacy, programming, automation design, writing, and problem-solving.
These answers are not wrong. But they have been repeated so often that they no longer bring the problem into focus.
So I changed the question.
What does AI do well, and where do people still matter?
What AI takes on
- Processing information at scale
- Finding patterns
- Producing first drafts
- Repetitive tasks
- Retrieving knowledge and applying rules
What stays with people
- Defining problems
- Setting goals
- Making value judgments
- Taking responsibility
- Building relationships and interpreting context
AI is getting better at helping with “how.”
But “what should we do?” remains a human decision.
The question shifts from capability to direction.
What is scarce changes with the times.
In the Industrial Revolution, labor was scarce.
In the internet era, information was scarce.
Now, information is becoming almost limitless.
Deciding what to ignore becomes harder.
AI can suggest a hundred strategies.
The challenge used to be finding a good answer.
Increasingly, it is choosing among abundant answers and taking responsibility for the choice.
AI does not remove a capability. It removes its scarcity.
Its scarcity has disappeared.
From this perspective, the question changes again.
Are we looking for capability only within the individual?
The original question assumes that, as the world changes, individuals must acquire new capabilities.
Yet historically, position has often mattered more than individual capability.
Across the Industrial Revolution, the internet, and mobile technology, the biggest differences came less from
“Who is smarter?”
and more from “What networks and assets can they build on?”
Competition happens between systems, not just people.
Many people frame this as “humans versus AI.”
One step further gives us “people using AI versus people without it.”
But the real comparison is:
versus organizations that incorporate AI
That makes individual capability only part of the picture.
Which organization you join, which networks you connect to, and which systems you build become more important.
Change the lens as often as you like. Some things remain.
Instead of searching for the essential skill, it may be more useful to ask what will remain scarce in the age of AI.
What is striking is that most of these things are not forms of knowledge.
They are closer to relationships, position, responsibility, and execution.
Which skills will be in demand in the age of AI?
What capabilities do I need?
What will remain scarce?