"The future of work" is usually presented at the wrong scale. Headlines discuss millions of jobs while you need to decide what to learn on Tuesday evening.

A 2020 World Economic Forum video introduced "The Great Reset" as a post-pandemic initiative focused on rebuilding economies to be fairer, more sustainable and more resilient. It was a policy and conference agenda, not a secret forecast of each person’s future.

Read forecasts as scenarios, not promises

The World Economic Forum’s 2025 employer survey estimates that labour-market change could create 170 million roles and displace 92 million by 2030 across the formal employment represented in its dataset. That is a net increase of 78 million, but the gain does not mean every worker or region benefits equally. These are employer expectations combined with labour data, not guaranteed counts.

The International Labour Organization’s 2025 GenAI index finds that one in four jobs worldwide has some potential exposure to generative AI. Its central conclusion is not that one in four jobs will disappear. Task transformation is considered more likely than full replacement, and exposure varies by occupation, income level, gender and actual technology adoption.

The OECD’s work on the net-zero transition makes a similar point. Many "green-driven" occupations are existing jobs with changing skills or rising demand, not entirely new green titles. That evidence primarily covers OECD economies, so it should not be copied directly onto South Africa. The principle remains useful: transitions often alter tasks inside familiar occupations.

Map your tasks before changing careers

Write down the ten tasks that occupy most of your current or recent work. Include unpaid projects if they show relevant ability.

Mark each task:

  • Growing: likely to become more useful.
  • Changing: still needed, but tools or standards are shifting.
  • Exposed: repetitive and increasingly easy to automate or reorganise.
  • Human-critical: depends on trust, judgement, physical context, accountability or care.

This is not a scientific automation score. It is a conversation starter.

The aim is not to predict whether your whole job survives. It is to identify the next useful skill.

Build a two-layer skill stack

Future-facing profiles combine one domain capability with one enabling capability.

Domain capabilities include customer service, nursing support, electrical work, logistics, finance, education or sales. Enabling capabilities include data literacy, digital tools, AI-assisted workflows, project coordination and sustainability knowledge.

Choose a pairing relevant to you, such as customer support with CRM reporting, electrical maintenance with solar fundamentals, or logistics with inventory data. Learning might begin with documentation, supervised exposure or a small project. Use an accredited programme where the occupation requires it and check local recognition before paying.

Move through an adjacent role

The most exciting job title may not be the safest first step. An adjacent move uses evidence you already possess while adding one new layer.

Ask:

  1. Which roles use at least 60% of my current strengths?
  2. What new tasks would they introduce?
  3. Can I test those tasks without resigning?
  4. Which evidence would an employer accept?

An administrative employee interested in data may first target reporting support rather than data science. A warehouse worker may explore inventory control before supply-chain analysis. An electrician may add renewable-energy exposure through recognised training and supervised work rather than claiming specialist status immediately.

Adjacency reduces the cost of being wrong.

Build proof before making a large bet

Create one small artefact related to the target:

  • A cleaned and explained spreadsheet.
  • A process map with an identified bottleneck.
  • A customer-response library.
  • A safe maintenance checklist based on approved procedures.
  • A short analysis of publicly available data.
  • A project summary showing decisions and lessons.

Remove confidential information and never simulate licences or regulated authority you do not hold.

Then show the work to someone who understands the field. Ask what feels realistic, what is missing and what entry-level employers actually test. One informed conversation can prevent months of unfocused study.

Use a 12-week career plan

Weeks 1–2: map tasks and research three adjacent roles.

Weeks 3–6: learn one defined skill and produce a small piece of evidence.

Weeks 7–8: obtain feedback and revise the work.

Weeks 9–10: update your CV and profile around verified evidence.

Weeks 11–12: submit targeted applications, request informational conversations and review the response.

At the end, decide whether to deepen, adjust or stop. Stopping a weak experiment is not failure; it is cheaper than building a career plan entirely from hype.

The future of work is not one event that happens to everybody at once. It is a series of changes in tools, tasks, investment, regulation and demand. You cannot control all of them. You can maintain a clearer map of your work, add useful capabilities and make the next move small enough to test.

Sources and limits

  • Forecasts have methodological and geographic limits. This article is an original practical synthesis, not a video transcript, political claim or guarantee of employment.