Beyond the Hype

AS a student looking ahead, it is remarkably easy to feel paralysed about the careers to come. We are constantly bombarded with headlines warning that artificial intelligence is coming for our jobs, that entry-level hiring is freezing, and that the traditional career path is disappearing. You put endless hours into studying, building your CV and planning your next steps, only to wonder whether your dream job will even exist by the time you graduate.

That anxiety is not irrational: the job market is shifting dramatically under our feet. But the real danger is not that AI will erase human labour entirely. It is that we are directing our fear at the wrong threat while failing to adapt how we learn and work.

What the Numbers Actually Say

It is true that the market is tight right now, especially for young workers. Eurostat put youth unemployment across the European Union at 15.5 per cent in June 2026, and across the OECD, the club of mostly rich economies, the equivalent rate has sat stubbornly around 11 to 12 per cent. The International Labour Organization’s most recent survey of youth employment, published in August 2026, found a global youth unemployment rate of 12.4 per cent, some sixty-seven million people, and a rise in the share of young people not in employment, education or training to a fifth of the age group.

A major factor driving this friction is indeed AI, but it is not replacing the entire workforce. It is disproportionately compressing the bottom rungs of the career ladder. Because AI tools can now instantly handle routine tasks like basic coding, data entry and drafting emails, companies simply do not need as many junior employees to do the initial grunt work.

The clearest evidence for this comes from the Stanford Digital Economy Lab, where Erik Brynjolfsson and his colleagues have been tracking payroll records for millions of American workers. They find that in the occupations most exposed to AI, meaning jobs where a large share of the day-to-day tasks are the kind a language model can now do, employment for workers aged twenty-two to twenty-five has fallen behind. The first version of their paper, in August 2025, put that shortfall at thirteen per cent. Their most recent revision, published in August 2026, puts it at roughly nineteen per cent.

Two things about that number matter enormously, and almost every headline gets them wrong. The first is that it is a relative figure, a gap rather than a count of jobs destroyed. It measures how far employment for young workers in exposed occupations has drifted below where it would have been had it kept pace with their peers in less exposed work. The second is that the authors are unusually blunt about what they have not found. They report no widespread, economy-wide job displacement linked to AI, and the adjustment they can see works mainly through firms hiring fewer young people rather than sacking the young people they already have.

That distinction is worth holding onto whenever you meet the kind of statistic that claims some enormous percentage of all jobs is about to be automated away. Those figures are almost always projections, estimates of what a technology could in principle do to a set of tasks, dressed up in the language of measurement.

The Stanford result is a different kind of thing: an observation about who is actually being hired. It is narrower, and far more reliable, precisely because it claims less. The ILO makes the same point from the other direction, estimating that only about six per cent of the jobs currently held by people aged fifteen to twenty-nine sit in occupations highly exposed to AI-related change. The squeeze is real, and it is concentrated. It is not a flood.

The Ticking Time Bomb

Automating routine tasks might look like a straightforward win for corporate efficiency, but freezing entry-level hiring creates a problem that only detonates later: a severe future talent shortage. When today’s managers and executives retire, firms will need experienced people to replace them, and those people have to start somewhere.

AI can certainly handle the grunt work that traditionally defines entry-level jobs, such as data processing, basic coding and formatting reports. What it cannot replicate is the experience, professional judgement and real-world problem-solving that young professionals develop while doing that work. Entry-level roles have historically been the training ground for future leaders. If firms drastically reduce or remove these positions today, they will inevitably face a leadership vacuum tomorrow.

This exposes a contradiction in the corporate world right now. Some companies are aggressively cutting junior roles to save money in the short term. Forward-thinking organisations recognise this as a trap: to have senior talent tomorrow, they must invest in training junior talent today.

PwC’s 2026 survey of more than a billion job advertisements across twenty-seven countries found that the companies most exposed to AI had grown their headcount by around half since 2018, against roughly a third at the least exposed firms. The firms getting the most out of the technology are using it to amplify what their people can do, not to shrink the payroll.

Role Redesign, Not Removal

Instead of eliminating junior talent, these firms are practising what might be called role redesign: keeping the early-career pipeline intact but shifting what a first job actually asks of you, away from repetitive data entry and towards managing the machine and thinking strategically.

The same PwC analysis puts numbers on this, and they are the most encouraging figures in the whole debate. Entry-level roles in AI-exposed sectors have grown by around thirty-five per cent since 2019, even as entry-level roles elsewhere in the economy have shrunk by about ten per cent. What has changed is not the existence of those jobs but their content: such roles are now roughly seven times more likely to demand the kind of human skills, such as leadership, judgement and creativity, that used to be asked of someone several years into their career.

That is a harder starting line, and it would be dishonest to pretend otherwise. But it is not a closing door. And stripping away the formatting and the data entry has a real payoff, because the skills that remain are the interesting ones.

The World Economic Forum’s most recent survey of global employers ranked analytical thinking first among the core skills companies need, with roughly seven in ten calling it essential, and creative thinking fourth; employers expect nearly two-fifths of what a worker needs to know to change by 2030. A successful young professional today is not a corporate drone, but someone who understands why things work rather than just executing repetitive steps.

To me, this is the dividing line between success and failure from here on. The people who make it will be those unafraid to ask basic questions, willing to dig deep into a topic, and able to use AI as a tool rather than a crutch.

I like to think of it like a calculator in a maths class. You can use a calculator any time you want, but if you just punch numbers into it without understanding the underlying principles, you will not get anywhere. The calculator handles the boring, repetitive arithmetic so you do not waste time, but you still have to do the actual thinking. AI works the same way. It takes care of the grunt work, but you are still the one driving the strategy.

Who Buys the Cars?

Beyond talent pipelines, there is a second reason it makes no economic sense for companies to stop hiring people altogether: they would destroy their own customer base. Businesses exist to sell things, whether to other companies or to ordinary consumers, and everything in an economy is connected. One firm’s wage bill is another firm’s revenue.

Imagine an extreme world in which AI takes every single job. To build cars, an automotive firm buys steel from a steel company and tyres from a tyre company, and uses those inputs to manufacture a vehicle. But if nobody has a job or a salary, nobody can buy the cars. The car company makes no revenue and loses money on all its raw materials. Because it cannot sell any cars, it stops buying steel and tyres, so those suppliers lose their customers too, triggering a domino effect down the entire supply chain.

That is an extreme scenario, and no serious economist expects it to arrive in full. But the underlying principle holds: the fewer people who have jobs, the less money is spent in the economy as a whole. When companies cut wages and jobs to save money, they end up shrinking their own market. Cutting costs can destroy your own sales.

Conclusion

Ultimately, the narrative that artificial intelligence will simply erase human labour misses the wood for the trees. The transition we are living through is not the end of the career ladder but a radical restructuring of it. As algorithms absorb routine tasks, the market is pushing us beyond being rote executors and towards becoming genuine problem-solvers. Companies that try to bypass human talent entirely will find themselves trapped between shrinking markets and leadership vacuums, because an economy cannot function without people to build, innovate and consume.

For our generation, this means the rules of the game are shifting, but the playing field is still wide open. We can build the necessary capability right now, in our everyday studies: treating AI as a study partner rather than an answer key, and using school projects to master underlying principles instead of just chasing grades. The path forward does not require us to out-compute the machine; it requires us to out-think, out-question and out-understand it. By trading busywork for genuine curiosity, we can step into a future where technology does not replace our potential, but unlocks it.