A baobab can live for more than a thousand years, its trunk wide enough to hollow into a room, its canopy visible for miles. What almost no one sees is what happens before any of that. In its earliest years, the tree builds a root system wider and deeper than anything above ground will suggest, while almost nothing visible signals progress.

A few weeks ago, I began a series exploring why Africa’s greatest advantage in the AI era may not be compute, but people.

The first essay used the honey badger to make a simple argument: Africa cannot win the AI race by trying to become a smaller version of the United States or China. We cannot match their capital, compute, or infrastructure at scale. Our opportunity lies elsewhere, in the capabilities we already possess and can develop further, particularly the human judgment that becomes more valuable as AI makes routine execution cheaper.

This essay moves further upstream.

If judgment is becoming more valuable, when does the capacity for judgment begin to form?

Why the causal chain matters

AI is making routine execution cheaper. As execution becomes cheaper, it becomes less differentiating. Anyone can now produce a competent first draft, a competent analysis, a competent summary.

What remains differentiating is the capability required to decide what should be executed at all, to judge whether the output is right, and to know when the confident answer is the wrong one. That capability is becoming more economically consequential, not because it has grown scarcer in some demographic sense, but because everything around it has grown more abundant.

Two categories of human capability matter here, and they are not the same thing.

Judgment: discernment, taste, ethical reasoning, interpretation, decision-making under uncertainty.

Relational agency: trust, empathy, accountability, collaboration, cultural fluency, deep listening.

Neither appears as a conventional subject on any syllabus. Both become more economically consequential as execution gets cheaper.

And this changes the question we should be asking about human capital.

If these are the capabilities we expect to matter more, where does the machinery that eventually produces them begin to develop?

Africa is already moving

Early childhood development has moved from a peripheral concern to a named continental priority. Africa’s new Continental Education Strategy for 2026 to 2035 places foundational learning at the centre of the agenda, with early childhood development forming part of that foundation. UNESCO’s January 2026 meeting in Dakar likewise framed early childhood education and development as strategic human capital infrastructure, not a marginal sub-sector.

The movement is visible at country level too. South Africa’s Department of Basic Education reported in its 2026 Budget Vote that ECD registration had grown by 200 percent between 2021 and 2026, with more than 1.2 million children now accessing registered ECD programmes. Rwanda’s national ECD plan raised daycare enrolment from roughly 1 percent in 2018 to 28 percent in 2022.

Africa is not starting from zero, and it does not need convincing that early childhood matters. That case has already been made.

What has not yet happened is connecting that investment explicitly to the capabilities an AI-saturated economy is beginning to value most.

Why the earliest window matters

The issue is not that African policy has failed to recognise early childhood. The issue is that even strong ECD systems can leave the earliest years and the people responsible for them insufficiently covered.

Research on care policy across the region finds that integrated ECD plans in countries including Rwanda, Kenya, Senegal and Côte d’Ivoire often leave children under three and their caregivers outside the main policy architecture. That matters because the earliest years are not simply a prelude to human-capital formation. They are part of the formation process itself.

James Heckman’s work offers a useful way to understand why.

The argument is not that judgment itself forms in infancy. It is that skills build on skills. Cognitive and non-cognitive capabilities develop together, and early investments can make later learning more productive. Heckman’s research on skill formation describes this process in terms of self-productivity and complementarity: capabilities developed at one stage affect the capacity to develop capabilities at later stages.

That changes how we should think about the education pipeline.

By the time a child reaches primary school, the system is not starting with an empty page.

The World Bank’s Learning Poverty measure illustrates the consequence. Before the pandemic, an estimated 86 percent of children in Sub-Saharan Africa were unable to read and understand a simple text by age 10. Post-pandemic modelling estimated that the figure had risen to 89 percent.

These are children already several years inside the formal education system.

If the foundations for attention, language, self-regulation, curiosity and learning are weak before they enter that system, later reforms are working against accumulated disadvantage.

We measure literacy. We measure numeracy. We are beginning to measure digital skills.

We barely measure the capacities that may become most valuable precisely when machines get better at execution.

What a three-year-old is actually building

A toddler is not developing judgment in any professional sense.

What is forming in those years are some of its precursors: self-regulation, curiosity, attention, language, social interpretation, emotional regulation, the capacity to negotiate, the tolerance to sit with frustration, the ability to distinguish one signal from another, the confidence to explore, the capacity to form relationships, and the willingness to act under uncertainty.

None of that is judgment.

All of it is part of the substrate on which judgment is later built.

While we should not project the capabilities of an adult professional backwards onto a three-year-old, we should ask what developmental capacities make those capabilities possible later.

And that matters in an AI era because, as machines take on more routine execution, human value increasingly shifts toward the ability to interpret, decide, collaborate, take responsibility and navigate uncertainty.

The foundation for those capabilities is being built long before anyone enters the workforce.

The reframing that matters

The most important shift this argument implies is not a new policy line item. It is a reframing.

Early childhood investment has traditionally been justified because children deserve healthy development, because school readiness improves later outcomes, and because childcare enables women’s economic participation. All of that remains true.

What has not yet been argued with the same conviction is that the foundational capacities human beings begin developing before they ever enter a classroom may become increasingly important in an economy where machines handle more of the routine work.

Africa’s continental frameworks now exist. What they have not yet done is make that connection explicit.

Two practical shifts follow.

Under-threes and their caregivers need to be brought inside national ECD plans as a matter of course, not treated as the population those plans still leave for later.

And ECD needs clearer, dedicated tracking in national financing, rather than being inferred from broader health and education expenditure. What is not measured is difficult to protect when budgets tighten.

Where this leaves the pipeline

Africa has started building the roots.

The question is whether we understand what those roots are ultimately meant to support.

By the time we call something a skill, much of the machinery that allows a person to acquire, apply and exercise that skill has already been developing for years.

A baobab does not become a baobab by tending only to its canopy.

The human-capital pipeline begins earlier than our workforce policies usually acknowledge.

And if Africa’s advantage in the AI era is going to be human capability, then the first question is not simply how we train better workers.

It is whether we are building the foundations from which better judgment can grow.