Artificial intelligence and the ageing workforce

By Matthew Parish, Associate Editor

Friday 5 June 2026

The first phase of the artificial intelligence revolution was often described as a youth revolution. The mythology surrounding Silicon Valley portrayed the young as the natural masters of the new technological order. University dropouts in hooded sweatshirts built billion-dollar companies from garages and dormitory rooms. Teenagers understood social media before their parents did. Coding was treated almost as a new adolescent language. Investors sought founders barely old enough to rent cars.

Yet as artificial intelligence moves from novelty into infrastructure, an unexpected phenomenon may be emerging. The labour market increasingly appears to reward older workers again. Not universally, certainly not absolutely, and not in every sector. Nevertheless there are mounting signs that the age structure of economic advantage may be tilting away from the very young and back towards people possessing long experience, institutional memory, and emotional stability.

This is a paradox because artificial intelligence is commonly assumed to accelerate generational displacement. Every previous technological revolution appeared to favour youth. Young factory workers adapted more readily to industrial machinery than ageing agricultural labourers. Young office workers learned computers faster than executives trained on paper filing systems. Social media culture rewarded rapid adaptation to constantly shifting digital environments.

Artificial intelligence may be different because it changes not merely the tools of work but the value of judgment itself.

Large language models, autonomous software agents and generative systems are extraordinarily effective at producing first drafts. They can write competent reports, analyse contracts, generate marketing materials, compose software code, and synthesise research in seconds. What they cannot yet reliably do is distinguish wisdom from plausible nonsense. Artificial intelligence systems imitate cognition without possessing lived experience. They predict likely answers statistically. They do not understand consequences in the human sense.

This distinction matters enormously in professional life.

For decades, many younger workers possessed a comparative advantage because they were quicker at acquiring technical skills. A 24-year-old programmer could often outperform a 55-year-old manager in manipulating digital systems. But when artificial intelligence automates much of the technical execution itself, the premium shifts. The crucial question is no longer “Can you produce output quickly?” but rather “Can you determine whether the output is correct, useful, lawful, ethical, strategically intelligent, and socially appropriate?”

These are questions that often favour older minds.

A senior lawyer reviewing an AI-generated contract may detect subtle risks because she has witnessed litigation disasters across thirty years of practice. A veteran journalist can identify when an AI-generated narrative “sounds true” but violates political reality or historical context. An experienced military officer may instantly perceive that an elegant AI-derived operational plan ignores morale, weather, corruption, or logistics. A seasoned financier may recognise that an apparently sophisticated investment thesis merely repackages an old fraud pattern.

Artificial intelligence therefore amplifies the economic value of accumulated pattern recognition.

This does not mean older people suddenly become better at using technology than younger people. Frequently the reverse remains true. Younger workers still adapt faster to interfaces, tools, and workflows. However the underlying economics may nevertheless favour older workers because artificial intelligence increasingly commodifies raw technical execution itself.

In earlier decades, junior workers justified their salaries partly through labour intensity. They produced drafts, analysed documents, compiled data, and executed repetitive professional tasks. Artificial intelligence now performs many of these entry-level functions rapidly and cheaply. This creates a dangerous structural problem for younger workers. If machines absorb apprenticeship tasks, then how do humans acquire apprenticeship experience?

Historically, professions depended upon gradual progression. Young employees learned through exposure to boring work. Junior lawyers reviewed documents endlessly. Junior journalists rewrote press releases. Junior bankers assembled spreadsheets through sleepless nights. Junior civil servants prepared briefing papers no minister would ever read carefully. Through these repetitive tasks, workers accumulated tacit knowledge.

Artificial intelligence threatens to remove the bottom rungs of this ladder.

Consequently firms may increasingly value workers who already possess mature judgement because the cost of training inexperienced staff rises sharply when machines handle the educational groundwork. A partner in a law firm may prefer one experienced associate supervising AI systems rather than six graduates learning the trade slowly. A newspaper editor may prefer veteran correspondents capable of rapid verification rather than armies of interns generating content already replicable by machines.

The result could be a hollowing out of the middle and lower tiers of white-collar employment.

This shift also intersects with demographic reality. Across much of Europe, North America, East Asia, and increasingly parts of Eastern Europe, societies are ageing rapidly. Labour shortages already exist in many sectors. Governments once assumed artificial intelligence would offset declining birth rates primarily through productivity gains. But artificial intelligence may simultaneously increase the economic importance of older cohorts by preserving their competitiveness for longer.

An experienced accountant who once struggled with software complexity may now use conversational AI interfaces effortlessly. A senior doctor may delegate paperwork and administrative burdens to AI systems while focusing on diagnosis and patient trust. A university professor can employ artificial intelligence to accelerate research preparation while relying upon decades of intellectual formation to interpret results critically.

Artificial intelligence thus acts partly as a compensatory technology for ageing cognition.

This is especially true because modern AI interfaces are linguistic rather than technical. Earlier digital revolutions required mastery of specialised software syntax. Artificial intelligence increasingly responds to ordinary language. Older workers who possess sophisticated verbal reasoning, negotiation ability, and conceptual understanding may therefore adapt more successfully than expected. Indeed many executives privately report that senior staff often use generative AI more effectively because they ask better questions.

Prompt engineering, in practice, frequently resembles management.

One must define objectives clearly, anticipate ambiguities, identify risks, refine outputs iteratively, and understand organisational context. These are executive skills more than youthful technical skills. The stereotypical twenty-year-old computer prodigy may know how to manipulate systems brilliantly, but a sixty-year-old diplomat may know far better what outcomes are actually desirable.

There is also a psychological dimension. Artificial intelligence creates immense informational instability. Vast quantities of text, imagery, analysis and propaganda can now be generated almost instantly. In such environments, emotional steadiness becomes economically valuable. Older workers often possess greater resilience amidst ambiguity because they have survived multiple economic cycles, technological transitions, and institutional crises already.

A generation raised during permanent digital acceleration sometimes assumes constant disruption is normal. Yet organisations under stress frequently seek precisely the opposite qualities: calmness, judgment, continuity, and credibility. Artificial intelligence intensifies uncertainty, thereby increasing demand for people perceived as reliable interpreters of chaotic systems.

This may partially explain why many senior executives appear unexpectedly enthusiastic about artificial intelligence adoption. Publicly they celebrate productivity gains. Privately they may also recognise that artificial intelligence strengthens hierarchical control within institutions. If AI systems reduce reliance upon large junior workforces, then organisations become more dependent upon smaller groups of experienced supervisors.

Such dynamics could reshape class structures profoundly.

For decades, higher education functioned partly as a mechanism for youth advancement into professional classes. Artificial intelligence threatens this bargain because credential acquisition alone no longer guarantees scarcity value. If a newly graduated analyst competes directly against AI-enhanced experienced professionals, the latter may dominate.

The consequences for younger people could become severe. Delayed career progression may translate into delayed family formation, reduced property ownership, and increasing intergenerational resentment. We may witness societies in which older workers remain economically active and professionally dominant far longer than previous generations, while younger workers struggle to establish themselves.

Yet caution is necessary before embracing simplistic conclusions.

Artificial intelligence will not uniformly favour older workers. In sectors requiring extreme adaptability, rapid experimentation, or deep computational innovation, youth may continue to dominate. Moreover many older workers still lack digital literacy entirely and may be displaced rapidly. The beneficiaries are unlikely to be “older people” generically, but rather experienced workers capable of integrating artificial intelligence into mature professional judgement.

Nor should one romanticise age. Experience can harden into rigidity. Institutions led exclusively by older cohorts often become stagnant, self-protective, and resistant to necessary change. Artificial intelligence may reward judgment, but judgment itself depends upon intellectual openness. A young worker capable of combining technological fluency with reflective maturity may prove more valuable than an older worker trapped by outdated assumptions.

Nevertheless the deeper historical point remains striking.

For perhaps the first time since the beginning of the digital age, technological transformation may not overwhelmingly privilege youth. Artificial intelligence reduces the scarcity value of technical execution while increasing the scarcity value of interpretation, trust, and responsibility. These are qualities often accumulated over time rather than learned instantly.

The artificial intelligence revolution therefore may not resemble the social mythology of Silicon Valley at all. Instead of replacing older generations with younger ones, it may create a strange synthesis in which machines perform the rapid cognitive labour once assigned to the young, while experienced humans increasingly occupy supervisory, interpretative, and strategic roles above them.

If so, then the defining economic conflict of the coming decades may not simply concern humans versus machines.

It may concern which humans remain economically indispensable once machines can imitate intelligence itself.