That agreement shaped how we designed our education systems and how organisations rewarded people. For many of us, it also shaped how we came to understand our own value.
And that is precisely the territory AI is now moving into. The International Labour Organization’s 2025 analysis of almost 30,000 occupational tasks found that one in four workers globally is in an occupation with some exposure to generative AI. Its conclusion was that jobs are more likely to be transformed than to disappear altogether, which matters. But this is not only routine work, either. Highly skilled, highly credentialed professional work is well within its reach.
What I am noticing is that underneath the more comfortable conversations about productivity and efficiency, many people are quietly carrying a question they rarely say out loud. If a machine can now do what I do, sometimes faster and occasionally better, what exactly remains mine?
We have been here before, though not quite in the way we often tell it
It would be easy to tell a neat story here. Machines took over physical work, so we moved into intellectual work. Now AI is taking over intellectual work, so we must move into something else. I have told a version of that story myself. The history, when you look at it closely, is more interesting than that.
Technology has never simply swapped one kind of human work for another. It has repeatedly reorganised the tasks that make up a job. Some tasks were automated, others were transformed, and entirely new kinds of work appeared that the previous generation could not have named. Computers were already doing cognitive work decades ago, in bookkeeping, calculation and administration, long before anyone spoke of generative AI. What they struggled with was anything that depended on judgment, context and genuine human interaction, and for a long time that boundary held.
That is the boundary now moving again.
I think it is important to be honest about what this history does and does not tell us. It does not promise an easy landing. When the boundary moved before, livelihoods were genuinely lost, whole trades disappeared, and the cost fell far more heavily on some communities than on others. Some of the unease people feel right now is not paranoia. It is an accurate perception of real risk.
What history does suggest, though, is that when technology has moved that boundary, it has not simply erased human value. It has relocated it, often slowly and at real cost, into forms of contribution nobody could fully see in advance. And each time, societies eventually changed what they intentionally developed in people.
Literacy expanded, then secondary schooling, then mass higher education, then technical and digital capability. What we expected a capable adult to know and be able to do shifted with the world around them.
The obvious response, and why I am not sure it is enough
So the natural question now is what we should be developing in people. The obvious answer is AI skills, and yes, we should. People will need knowledge, judgment and expertise to use increasingly capable technology well. Intellectual capability is not becoming irrelevant.
My sense, though, is that intellectual capability on its own may be becoming an increasingly incomplete answer to what we should be developing in human beings.
I want to be careful about how I make this argument, because there is a tempting version of it that I think is wrong. It goes something like this: here is a list of things AI cannot do, so those must be what makes us human. The difficulty is that any such list comes with a shelf life. Technology has a long habit of eventually doing what we confidently said it never would. If our sense of worth depends on the boundary staying exactly where it is today, we will need to keep redrawing it, and it will keep moving.
So I am less interested in what machines cannot do, and more interested in what human beings need in order to live well.
For generations we have invested enormously in developing the intellect. We have invested far less intentionally in the capacities that allow people to navigate emotion and difficulty, build and repair relationships, make meaning when life disrupts the story they were telling themselves, take ownership of what is theirs to do, and show up for one another when it matters. Most of us were never explicitly taught these things. We simply assumed people would pick them up along the way.
These capacities do not matter because AI happens to be weak at them. They matter because they are part of how human beings flourish, and they would matter even if technology became far more capable than it is today. That, I think, is what makes this moment an opportunity as well as a threat. Not a reason to compete with machines, but a reason to become more intentional about what we develop in the human being who will be using them.
The question underneath the question
This is my own thinking rather than something the research has settled, and I hold it with some humility. History supports the need to reconsider what we develop in people when technology changes the ground beneath us. It does not tell us what we should develop next. That part is ours to work out.
But I am increasingly wondering whether the most important question of this moment is not really about Artificial Intelligence at all, however much it may look like one.
Perhaps the question is not what AI will replace. Perhaps it is what AI is inviting humanity to become.
And I suspect the answer has far less to do with machines than it does with us.
Rentia
Reference
Gmyrek, P., et al. (2025). Generative AI and jobs: A refined global index of occupational exposure (ILO Working Paper 140). International Labour Organization and NASK. https://webapps.ilo.org/static/english/intserv/working-papers/wp140/index.html