Last year the idea of AI replacing jobs stopped feeling like a news story and started feeling personal. I had been aware of the economic shifts for some time: accountancy firms cutting trainee intakes; law firms automating junior research. But now the problem had arrived at the dinner table.
My 16-year-old daughter Thea had her sights set on a career in finance and was passionate about it. Yet I kept thinking about the entry points of a finance career and how they were the roles AI was absorbing fastest. Having spent 25 years at Microsoft leading AI and customer experience initiatives, I knew only too well how disruptive this technology could be and I wondered whether the path Thea was aiming for would look the same by the time she arrived. I wanted to advise, but needed to find some answers first.
My first instinct was to look at the data. I set about collecting intelligence, cross-referencing research and looking for patterns. Evenings. Weekends. Seventeen sources: the World Economic Forum (WEF), Goldman Sachs, McKinsey, PwC, Stanford HAI, Anthropic and others. I was not trying to build something reassuring for Thea. I was trying to build something honest. The result was a 20-page document to help guide young people through the upheaval of AI. This is what I found.
The disrupted ladder
The most important thing to understand is also the most counterintuitive. AI is not demolishing careers from the top. It is removing the bottom rungs of the ladder first.
Think about how almost every professional career develops. A “Big Four” trainee at the accountancy firms Deloitte, KPMG, EY or PwC reconciles spreadsheets and drafts standard documents. A junior solicitor reviews contracts. A graduate analyst builds financial models. These are the apprenticeship stages. They are how young professionals develop the judgment that eventually makes them irreplaceable.
AI performs many of those tasks faster and more cheaply. KPMG has cut its UK graduate intake by 29 per cent, Deloitte by 18 per cent, EY by 11 per cent. Graduate job adverts in accountancy are down 44 per cent year-on-year. UK entry-level postings have decreased by nearly a third since ChatGPT appeared.
For the first time in modern history, a generation may graduate without a clear way to become experienced. That is not just a jobs story. It is a human development story.
But here is what every doom headline misses. The WEF projects 170 million jobs created by 2030 alongside the ones displaced. The net picture is not collapse. It is transformation. And transformations reward the families who understand them early.
The superpowers you need
When I worked through the research, the same four human capabilities appeared across every source as the ones AI cannot replicate. Emotional intelligence. Creative vision. Physical dexterity. Ethical judgment.
But identifying AI-resistant qualities is only half the picture. Today’s young people also need to be fluent in AI itself. PwC tracked the financial value of this most precisely. Workers with genuine AI fluency earn 56 per cent more than those without it, in every field. That wage premium doubled in just 12 months. AI-related job postings are growing three times faster than average, and the gap is widening every year.
Think back to the 1990s. Teaching children to type was not just teaching a skill. It gave children a passport to professional life. AI literacy is that same transition, arriving faster and cutting deeper. Learning to use AI fluently in 2026 is what learning to type was in 1995.
Parenting newsletter
Join our community receiving professional advice and sharing personal stories.
Sign up with one click
For parents wondering where to start: Khan Academy’s Khanmigo tutor is available to pupils of any age. Google’s AI Essentials course requires no technical background. For older teenagers, the coding tool GitHub Copilot is free for students and teaches AI-assisted coding through practice. The goal is not to produce programmers. It is to produce young people who understand what these tools can and cannot do, and that understanding is worth more than almost any specific qualification.
Careers that will survive (and won’t)
To establish which careers will thrive in the AI age, I set up a framework that scored 35 of them across nine categories including emotional intelligence, creative thinking and vulnerability to AI tools. A score closer to 100 per cent means the role depends heavily on things AI cannot replicate. A score closer to 35 per cent means much of the work is already within reach of automation. Here are the winners and losers, by sector.
Healthcare scores highest of any sector. Surgeons, GPs and consultants score 91 to 96 per cent on long-term resistance. The diagnostic complexity, physical skill and ethical judgment required in medicine cannot be replicated by any system in development. The NHS faces a structural workforce crisis: it needs more than 50,000 more nurses, mental health practitioners in particular, as waiting lists lengthen. Mental health counsellors score 98 per cent. No chatbot replaces a therapeutic relationship built over months with a person in genuine distress. The sector is projected to add close to a million roles in the UK over the next 20 years. A newly qualified NHS nurse starts at around £32,000 and can progress to over £50,000 in specialist roles. Job security of this kind is something financial services can no longer promise.
Education scores 78 per cent to 88 per cent. AI can deliver information. It cannot notice that a child is struggling at home or adapt in real time to a difficult moment in a classroom. Teacher shortages in Stem subjects and special educational needs are acute, and the government offers training grants and financial incentives to attract graduates.
Skilled trades score between 82 and 94 per cent. Electricians average about £47,000 annually, and experienced tradespeople earn considerably more. There are 227 job openings for every apprenticeship place, and the country needs 104,000 new electricians by 2032 just to meet housebuilding and renewable energy targets. Plumbers and construction managers sit in similar territory. These are structural shortages of jobs that require physical judgment in novel environments. A leaking pipe in a Victorian terrace is not a repeatable task, and no robot can handle it.
Law scores differently depending on where you are within it. Partners and barristers who exercise judgment, advocate for clients and navigate complex human situations score 74 to 82 per cent. Paralegals and junior associates doing document review and contract analysis score 35 to 45 per cent. The profession is not disappearing, but young people considering law need to think carefully about whether the skills they are building will still be in demand by the time they qualify.
Finance divides similarly. Senior fund managers, financial advisers building long-term client relationships and risk managers exercising judgment score 68 to 78 per cent. Junior analysts doing data processing and model building score 38 to 50 per cent and falling.
Technology tells a nuanced story. Pure software developers score 62 per cent; which is falling as AI-generated code improves rapidly. Cybersecurity professionals score 79 to 85 per cent, since the adversarial nature of the field means humans must stay ahead of threats that are themselves AI-powered. Cloud architects and AI engineers score higher still. UK AI engineers average above £68,000.
The creative industries score 71 to 84 per cent for roles involving genuine original thought, cultural judgment and human perspective. AI generates images and text at scale, but it cannot come up with ideas rooted in experience or edit with the instinct that comes from decades in a field. The roles most at risk are those involving high-volume, templated production rather than genuine creative leadership.
Diplomacy, international relations and public policy score 88 to 95 per cent. Negotiating between sovereign nations, building coalitions across cultural divides and making decisions under geopolitical pressure is precisely the kind of complex, trust-dependent, ethically freighted work that sits furthest from what AI can do.
The careers that score lowest are where tasks are most predictable and repeatable: paralegal at 35 per cent, accountant at 45 per cent, data entry and administrative roles at 15 to 25 per cent. This does not mean those careers are finished. It means the automatable parts will be automated and the parts requiring human judgment will become more valuable.
The new careers
The most valuable of the new jobs that the WEF predicts will exist by 2030 will sit in the intersections between disciplines. The biomedical data scientist who designs clinical trials and builds the AI pipeline to interpret them. The clean energy engineer modelling grid distribution and solar deployment for the UK’s net-zero targets. The AI policy specialist ensuring automated decisions treat people fairly, a role that barely existed three years ago and is now being recruited for in governments and regulators.
These roles have something in common. They require deep knowledge in a substantive field combined with genuine AI fluency. Neither alone is sufficient. A young person who studies international relations and learns to use AI analytical tools is not choosing between two things. They are building the combination that will be hardest to find and most in demand.
What to do now
Chona Kasinger for The sunday Times
After many dinner table conversations, Thea said something that stopped me cold: “Diplomacy cannot be outsourced to robots, Dad.” Instead of finance she chose to do international relations: learning how humans negotiate, build trust and resolve conflict. Every one of those skills is in the 90th percentile for AI resistance.
We are facing a particular moment in history. It is not one that will announce itself. There is no letter from school, no official notification that the world your child is preparing for has quietly become a different one.
The families who will look back on this decade without regret are the ones who had the conversation early and trusted that a child who understands the world they are entering is far better equipped than one protected from it. Here are some first steps:
If your child is 10 to 12, build the foundations: teach them to be curious by reading carefully and arguing a point. Curiosity is the hardest quality to automate.
If your child is 13 to 15, have one conversation this week. Not a lecture. Ask what they think AI is doing to the world. Help them begin using AI tools, not to do homework for them, but to understand what these systems cannot do. That understanding is the first superpower.
If your child is 16 to 18 and making real choices, look hard at where the four human superpowers appear in the careers they are considering. AI fluency is not optional any more. The wage premium for those who have it is visible and growing fast. “Wait and see” is not a neutral position. It is a decision. The data says it is the wrong one.
Babith Bhoopalan is founder of the consultancy Quantumleap Insights and creator of the AI Career Playbook, which includes a curated toolkit of resources for children, and the AI Career Playbook Generator, which will produce a personalised action plan