Which South African Jobs Should Never Be Automated? Bill Gates Has Started the Argument

His idea of "Human Reserved" work deserves a serious local answer, because the lines will be drawn differently here than in Seattle or Tokyo.

Sign up for my Substack daily AI newsletter here.

See my AI Training course portfolio for corporate Business Leaders here.

Article link: https://johanosteyn.substack.com/p/which-south-african-jobs-should-never?r=73gqa&utm_campaign=post&utm_medium=web&showWelcomeOnShare=true

Follow me on LinkedIn: https://www.linkedin.com/in/johanosteyn/

Bill Gates has spent most of his public life as artificial intelligence’s most credentialed optimist. In an essay published on 26 August, titled The Turbulent AI Era Is Here, he changed register. The essay is worth reading in full, but the proposal I want South African readers to notice is the one that has drawn the least local attention. Gates argues that some categories of work should be designated as Human Reserved, roles society deliberately chooses not to automate even when machines become technically capable of performing them. He borrows the image from nature reserves: places where roads and buildings could be built, but are not, because the loss would be too great.

CONTEXT AND BACKGROUND

Gates grounds the proposal in his father’s death from Alzheimer’s disease in 2020. His father was cared for day and night by paid caregivers who understood him even when he struggled to express himself. He could not always say when he was hungry, but they always knew. Something in that care, Gates writes, was irreplaceably human. He extends the same reasoning to work that carries hard news, work that demands moral judgement under pressure, and work whose loss would displace people who cannot easily change occupations. The lines, he says, will vary by country. Japan, with a shrinking workforce and too few young people to care for its elderly, may welcome caregiving robots that another country would refuse.

The essay identifies employment disruption as the first of three risks that world leaders are unprepared for, alongside more powerful bad actors and the effects on children and human relationships. The Stanford Digital Economy Lab has found that after the widespread adoption of generative AI, employment fell significantly among young workers in occupations most exposed to replacement, but not among their older colleagues. That is the pattern Gates expects to widen. Human Reserved is his answer to it.

INSIGHT AND ANALYSIS

The proposal matters because it reframes the question. Most South African conversations about AI and work are stuck in a tired binary. Will AI take our jobs, or will it not? Gates has quietly replaced that with a better question. Which jobs do we want to keep, and why? The moment the question changes shape, the answer becomes a matter of choice rather than forecast, and South Africa has choices to make that Seattle and Tokyo do not.

Consider what a South African Human Reserved list might contain. Frontline health workers in a public system that has too few of them already. Teachers in classrooms that routinely hold forty children. Social workers in a country that has the largest population of children orphaned by AIDS in the world. Home affairs officials who can look a person in the eye when the birth certificate is finally issued. CCMA commissioners resolving disputes that carry the weight of livelihoods behind them. Court interpreters working across South Africa’s official languages. The caregivers who look after the elderly in townships and villages where formal frail care does not reach. These are not roles we should reserve because AI could not eventually assist with parts of them. The distinction that matters is between augmentation and replacement. A social worker whose case notes are drafted by AI is being helped to do her job. A grieving family receiving a terminal diagnosis from a chatbot is being denied something Ubuntu tells us they were owed.

I have previously written about the pattern of organisations measuring AI-era productivity while failing to measure whether their people are still learning, and the same warning applies to work whose value is not captured by productivity metrics at all. A caregiver who knows when a patient is hungry without being told is doing work no dashboard will register. So is a teacher who notices that a child is coming to school without breakfast. Automating that work does not eliminate it. It moves it onto the shoulders of the family, the neighbour, or nobody at all.

Gates is honest that the proposal raises questions he cannot answer. Who decides what is reserved, by what criteria, how do you keep companies from replacing people with machines regardless, and what happens to international trade when one country permits automation of a task that another has reserved? These are hard questions. They are also questions South Africa is unusually well placed to answer, because we already have institutional experience of drawing similar lines in labour law, in professional licensing, and in Broad-Based Black Economic Empowerment. Whatever one thinks of those instruments, they demonstrate that a democratic society can decide, in public, that some economic outcomes matter enough to shape markets around them.

IMPLICATIONS

For South African employers, the practical implication arrives sooner than the legislative one. Any organisation making adoption decisions this year is already drawing an implicit Human Reserved line, just without naming it. Every call centre decision to route the first minute of a difficult conversation to a person rather than a bot is such a line. Every hospital decision to keep a nurse in a room where a triage algorithm would suffice is another. The choice is not whether to draw the line. It is whether to draw it deliberately, with a named human accountable for the decision, or to let it be drawn by whichever vendor’s default settings the procurement team accepted. In any process where a machine contributes to a decision that matters, a human being has to remain a net contributor to that decision, and accountability has to sit with a named person who can explain and defend it. Human Reserved work is that principle applied at the level of an economy rather than a workflow.

CLOSING TAKEAWAY

Gates has given the world a new category. The countries that fill it in first will shape the ones that come later. South Africa has both the reason and the standing to be among those countries, because our labour market, our care burden and our unemployment rate mean the cost of getting this wrong is measured in lives rather than in quarterly earnings. The argument to have, in cabinet, in boardrooms, in universities and in union offices, is not whether Human Reserved work is a good idea. Gates has already made that case. The argument to have is which work belongs on our list, who decides, and by when. That conversation should start now, because the lines are being drawn regardless of whether we are in the room.

Johan Steyn is a prominent AI thought leader, speaker, and author with a deep understanding of artificial intelligence’s impact on business and society. He is passionate about ethical AI development and its role in shaping a better future. Find out more about Johan’s work at https://www.aiforbusiness.net