by Arabinrin Aderonke

Artificial intelligence has moved from the pages of science fiction into our homes, offices, schools, hospitals and businesses. It can write articles, analyse documents, create images, translate languages, write computer code, summarise meetings and assist professionals with tasks that once required hours of human effort. As AI becomes increasingly sophisticated, a question is beginning to concern parents, young people and professionals everywhere: what will happen to human work over the next 10 to 20 years?

The answer may be more complicated than the popular fear that AI will simply take everybody’s jobs. Increasingly, the evidence suggests that AI is more likely to change the nature of work than eliminate entire professions. The World Economic Forum estimates that significant job creation and displacement will occur by 2030, while research from organisations such as the International Labour Organization and PwC suggests that many occupations will be reorganised as AI takes over some tasks and increases the value of others.

This means that the right question may not be, “Which jobs will AI destroy?” A better question is, “Which parts of our jobs will AI do, and which parts will still require human beings?” That distinction is critical because a machine may be capable of performing 50 or 70 per cent of the tasks associated with a profession without eliminating the profession itself. Human beings may still be required to supervise the technology, interact with clients, make difficult decisions, handle unexpected situations and accept responsibility for the outcome.

There is little doubt that some categories of work will be heavily affected. Jobs involving repetitive, predictable and highly structured tasks are particularly exposed. Data entry, routine bookkeeping, basic administrative work, transcription, simple customer service, document processing and some forms of content production are examples. AI systems can process enormous quantities of information at a speed no human can match, and businesses will naturally use that capability to reduce repetitive work and improve productivity.

Accounting provides a good example. AI can already process transactions, identify inconsistencies, organise financial information and assist with routine reporting. Over the next decade, much of the basic bookkeeping that currently requires human labour could become automated. This does not necessarily mean that accountants will disappear. Instead, the accountant may move away from entering figures and spend more time interpreting information, advising clients, investigating irregularities and making professional judgments. The World Economic Forum already identifies accountants and auditors among roles expected to face pressure from technological change.

The legal profession is likely to experience a similar transformation. AI can search thousands of documents, identify relevant cases, compare contracts and summarise large volumes of legal information within seconds. Routine legal research and documentation could therefore become much faster and require fewer people. But the practice of law involves more than information retrieval. Lawyers negotiate, persuade, strategise, interpret human behaviour and represent clients in situations where the consequences can be enormous. A machine may help a lawyer prepare a case, but society may still insist that a human professional stands behind the advice and takes responsibility for it.

Journalism will also change significantly. AI can write routine news reports, generate headlines, transcribe interviews and produce social media content. But journalism at its highest level is not simply writing. Investigative journalism requires building relationships with sources, earning the trust of whistleblowers, travelling to places, asking difficult questions, verifying information and sometimes confronting powerful people. A machine can help a journalist analyse documents, but it cannot easily replace the human courage, relationships and judgment involved in discovering a story in the first place.

Programming is another profession likely to experience major transformation. AI can already generate code and assist developers in solving technical problems. Over the next 10 to 20 years, routine coding may require considerably fewer hours of human labour. However, software development involves understanding what should be built, why it should be built, how it should interact with people and what risks it might create. The programmer of the future may therefore spend less time writing every line of code and more time designing systems, supervising AI-generated code and solving complex problems.

The more interesting question concerns professions where AI may be powerful but human presence remains fundamental. Healthcare is perhaps the clearest example. AI will become increasingly capable of analysing medical images, identifying patterns in patient data and assisting doctors with diagnosis. Robots will also perform increasingly sophisticated physical tasks. Nevertheless, healthcare involves a human body, a human mind and often a frightened family. Nurses, doctors, physiotherapists and caregivers do not simply process information. They touch patients, observe them, reassure them, respond to unexpected changes and make decisions under pressure.

Nursing and caregiving may therefore prove particularly resilient. A machine can remind an elderly patient to take medication, monitor vital signs or assist with movement. What it cannot easily replace is the comfort of another human being sitting beside a frightened patient, holding a hand or recognising emotional distress. As populations age, these human elements of healthcare may become even more important.

The same principle applies to physiotherapists, occupational therapists, dentists and many other hands-on medical professionals. Their work involves physical examination, fine motor skills and constant adaptation to individual patients. Even if robotics becomes extremely advanced, the unpredictable nature of the human body means that professionals will still need to make decisions when reality does not behave exactly as expected.

Skilled trades may also be more resilient than many people realise. Electricians, plumbers, mechanics, carpenters and other technicians work in environments that are rarely perfectly predictable. A plumber may arrive at an old building and discover that the pipes are nothing like the drawings. An electrician may encounter an installation that was badly modified years earlier. A mechanic may hear a sound that does not correspond neatly to a diagnostic code. These professions require physical dexterity, improvisation and the ability to solve problems in the real world.

Construction will certainly become more automated. Drones, robotics, AI-assisted design and autonomous machinery will transform the industry. Yet construction sites remain dynamic environments involving weather, terrain, materials, machinery and dozens of workers interacting simultaneously. Humans will continue to be required for many forms of skilled physical work, supervision and problem-solving, even as machines take over more repetitive activities.

Emergency professions may be even more difficult to automate completely. Firefighters, paramedics and disaster-response workers operate in chaotic environments where conditions can change within seconds. A firefighter may have to make a decision that was never anticipated in a computer model. A paramedic may have to comfort a frightened person while simultaneously assessing a medical emergency. AI can provide information and predictions, but a human being may still have to enter the building, treat the patient or make the final decision.

Education presents another interesting case. AI can explain mathematics, generate lesson plans, answer questions and personalise learning materials. It could become one of the most powerful educational assistants ever created. But teaching is not simply the transmission of information. A good teacher recognises when a child has lost confidence, motivates a struggling student and helps young people develop discipline, curiosity and character. For early-childhood education and special-needs education in particular, the human relationship may remain indispensable.

Psychology, counselling and social work are similarly dependent on human connection. AI may become remarkably sophisticated at conversation and emotional analysis, but many people will continue to want a human being when discussing grief, trauma, family problems, fear and major life decisions. Trust is not merely a technical function. The relationship between two human beings is itself part of what makes many forms of care effective.

Leadership may ultimately become one of the most valuable human capabilities in an AI-driven economy. AI can analyse information, identify patterns and recommend strategies, but leadership requires persuading people to follow a vision. It requires courage, negotiation, emotional intelligence, political judgment and accountability. A chief executive, political leader, community leader or entrepreneur may have an army of AI systems providing analysis, but at the end of the day people will still ask: Who made the decision? Who is responsible? Who do we trust?

This is why the future may reward people who combine technological competence with deeply human abilities. Current research is already pointing in this direction. PwC’s 2026 AI Jobs Barometer found that AI-exposed roles are increasingly demanding traditionally human-intensive skills such as leadership, creativity and judgment, while the International Labour Organization similarly identifies adaptability, resilience, human agency and socioemotional skills as increasingly important.

The great divide, therefore, may not be between “AI jobs” and “human jobs.” It may be between predictable work and unpredictable work; routine work and judgment-based work; isolated digital tasks and human relationships. The more a job depends on repetitive information processing, the greater the potential for automation. The more it depends on trust, physical presence, complex judgment, emotional intelligence, leadership and unpredictable environments, the more difficult it becomes to automate completely.

This does not mean that anyone should choose a career simply because it appears “safe from AI.” Technology will continue to evolve, and robotics could eventually transform many physical professions as well. Instead, young people should be prepared to become exceptionally good at something AI cannot easily reproduce: understanding people, solving difficult problems, building relationships, exercising judgment and taking responsibility.

The safest career of the future may therefore not be one that avoids AI. It may be one in which a human being knows how to use AI better than everyone else while retaining skills that machines cannot easily replicate. A doctor who uses AI intelligently may outperform a doctor who refuses to use it. A journalist who combines AI-assisted research with exceptional investigative instincts may become more productive. A lawyer who uses AI for research but possesses extraordinary negotiation skills may become more valuable. A teacher who uses AI to personalise learning while providing genuine mentorship may become indispensable.

The future of work may not ultimately be human versus machine. It may be human with machine.

AI will increasingly provide speed, scale, memory, pattern recognition and automation. Humans will continue to provide judgment, empathy, leadership, courage, creativity, relationships, physical intervention and accountability.

The children entering school today may therefore enter a labour market very different from the one their parents knew. Their greatest advantage may not be knowing everything. Machines will increasingly know more than any individual human can know.

Their advantage will be knowing what to do with that knowledge, how to work with intelligent machines, how to work with other human beings, and when a human being—not a machine—must make the final call.

That may be the real career lesson for the next 20 years: do not prepare the next generation to compete with AI. Prepare them to become the humans AI cannot replace.