AI.

I’m seeing this repeatedly in my conversations recently: we still think in terms of the previous ten years.

Much of the technology we adopted replaced something in our lives. A workflow, a notion, a way of thinking. It empowered us to do things we couldn’t do before.

Over time, linear thinking became a convention. A task requires ten people. A technology reduces that to five. We extend the same line and imagine it eventually requiring nobody.

But we keep the task itself fixed. We count how much work disappears without asking what becomes possible when those ten people can do more.

Our mistake nowadays: we look at AI mainly through what it replaces.

We know it is massive. Yet we look at today’s work, imagine AI replacing every piece of it, and call that the future.

The anxiety around it is understandable, but I think the certainty behind this anxiety is deeply premature.

AI has already augmented what we can do beyond what many of us imagined. Yes, companies can cut jobs quickly, sometimes before they have figured out what they can actually replace. But how much opportunity has opened up in the same period?

In many parts of our work, we are simply, plainly more capable with AI.

For sure, AI can make certain jobs redundant. Making work less necessary has been part of technology’s promise since we first put oxen to work pulling ploughs. Much of our investment in technology relies on that promise.

On the other hand, how many more people can now access knowledge they couldn’t before?

How much can AI improve our food system? Our health? Our general understanding of life?

How much opportunity does it create for small business owners? How much of its potential is yet to be realized?

We talk about knowledge work disappearing. Losing your income is a real consequence. A future opportunity does not pay today’s bills.

But what happens to the person who can now build and run something that previously required a team, capital and years of learning?

What do we call this new class of self-employed small business owners? Are they freelancers? Founders? Generalists? Something we haven’t properly understood yet?

Being able to build something does not automatically mean being able to make a living from it. More people being able to build also means more competition. But it changes who gets to try.

I’m not saying teams, capital or years of learning have become redundant. I’m saying those requirements shaped how we think about who can build something valuable. AI is changing some of those requirements, and our thinking hasn’t caught up.

We don’t even have a settled description of what is emerging, yet we speak so confidently about where this ends.

It is easy to name many of the negative effects because we recognize what is being lost. Those are our known knowns.

But what are the unknown unknowns? What becomes possible when more people can learn, experiment and build at a much lower cost?

AI is already helping researchers investigate molecular structures and interactions. It is helping automate repetitive, uncreative, boring processes.

Yes, AI’s water consumption deserves scrutiny. So do the water savings it enables in agriculture and other industries. Counting only one side gives us an incomplete picture.

I understand: none of this proves every benefit will outweigh every cost. But neither does counting the work it replaces prove that our future is less bright.

From electricity to cars, from the plough to the steam engine, new technologies have replaced work and created new kinds of work. They gave us new things to learn, adapt to and build upon.

That history gives us reasons to look beyond replacement. It gives us no guarantee that the transition will be fair or painless.

With AI, many of us were caught unprepared.

We have to be faster in understanding its capabilities. Faster in learning how to use it carefully. Faster in preparing for the consequences, including for the people whose work changes before they have somewhere else to go.

That takes diligence, judgment and an openness to learning the new world as it emerges.

And this is why I think our picture of AI is still too small.

We are trying to understand an expansion of human capability by counting the pieces of today’s work that disappear.

Anyway, thought to drop my €0.02 on the subject. What do you think?