When Skills Expire Before Careers Do
Technology is shortening the useful life of knowledge. The harder question is who gets the time, money and institutional support to keep up.
A career can last forty years. Some of the knowledge that begins it may not last ten.
For most of modern history, we built education around a simple sequence: learn first, work later. That sequence is beginning to collapse. Technology is changing occupations while people are still performing them, turning learning from a phase of life into something closer to permanent professional maintenance.
The question is no longer simply whether people should keep learning. It is who carries the burden when learning becomes a permanent condition of working life.
The Old Bargain: Learn First, Work Later
The twentieth-century version of this bargain was straightforward. People spent years acquiring a qualification, entered a profession and expected experience to deepen what they already knew. Further training existed, but it was usually supplementary. The basic assumption was that the knowledge carried into work would remain useful for a long time.
Artificial intelligence, automation, digitalisation, new regulation, changing supply chains and the green transition are altering occupations while people are still performing them. The result is not that degrees have become obsolete. It is that initial education can no longer be expected to carry a person through an entire career.
Cedefop estimates that around six in ten European employees are susceptible to some degree of AI-related task transformation, making continuing skills development increasingly important rather than exceptional.
Knowledge has a maintenance cost.
Europe’s Uneven Learning Landscape
Under the European Pillar of Social Rights Action Plan, the EU has set a target for at least 60% of adults to participate in training every year by 2030. Lifelong learning is also explicitly recognised within the European Pillar of Social Rights.
Yet the reality remains uneven. In 2025, 13.7% of EU adults aged 25–64 reported participation in formal or non-formal education or training during the previous four weeks. Sweden reached 38.2%, Denmark 31.0% and Finland 28.1%. At the other end of the distribution, Greece stood at 5.2%, Bulgaria at 6.1% and Croatia at 6.5% (Eurostat, 2026).
There is an important methodological point here. The Eurostat indicator asks about learning in the previous four weeks. OECD figures using a 12-month window naturally produce much higher participation rates: around two in five adults across participating OECD countries reported formal or non-formal job-related learning in the 2023 Survey of Adult Skills. These figures should not be compared as though they measured precisely the same thing.
What they do show together is that learning has moved well beyond the university campus. In OECD countries, only about 8% of adults participate in formal education in a year, while roughly 37% participate in non-formal job-related learning. The modern learning system is increasingly made up of workplace courses, certifications, digital modules, short programmes, professional retraining and learning embedded inside employment itself.
The Lifelong Learning Paradox
The central paradox of lifelong learning is that the people who may have the greatest need to update their skills are not necessarily those with the easiest access to further learning. OECD evidence shows that participation among low-skilled adults is substantially lower than among medium- and high-skilled adults, with an average gap of more than 30 percentage points. Older workers, people in precarious employment and other disadvantaged groups are also under-represented in adult learning systems. Unless access to learning follows need rather than existing advantage, lifelong learning risks reinforcing some of the inequalities it is supposed to overcome.
This becomes clearer when “reskilling” moves from policy language into everyday life. A professional in a large organisation may be offered paid courses, certifications, digital tools and protected learning time. Someone working irregular hours in a small business may be expected to acquire comparable skills after work, at personal cost and while managing family responsibilities. OECD data show how concrete these constraints are: among adults who wanted to participate in learning but encountered an obstacle, 48% cited lack of time because of work or family responsibilities, 14% reported the absence of suitable training and 13% pointed to cost. The problem, then, is not simply whether adults are willing to learn. It is whether the conditions of their lives actually allow them to do so. And that leads to a more difficult question: who pays for remaining employable?
The Hidden Learning Tax
When an employer introduces a new technology and provides training during working hours, the cost of adaptation is largely institutionalised. The organisation absorbs not only the direct expense of training, but also the time required for people to become competent again. When workers are instead told to “stay relevant”, subscribe to courses, learn new software or gain new certifications in their own time, that cost quietly shifts from the institution to the individual. This is not a tax in the fiscal sense. It is a hidden learning tax: the private cost workers increasingly pay simply to maintain their professional relevance.
But that tax is paid in more than money. It is paid in evenings, weekends, family time and attention. For one worker, adaptation may be part of the job, supported by an employer and completed during paid working hours. For another, the same process becomes a second job performed after the first one ends. The difference is not simply a matter of willingness to learn; it is a question of who absorbs the burden of remaining professionally current.
If continuous learning becomes a permanent requirement of working life, it should not be designed as an endless personal examination in which workers are expected to prove, repeatedly and at their own expense, that they still deserve to remain employable. The objective is not to create adults who permanently feel underqualified, but to build systems in which adaptation becomes a normal part of working life itself. Time for learning, access to training and financial support should therefore be understood not simply as occasional benefits, but as part of the infrastructure required for technological change.
Get Markos G.’s stories in your inbox
Join Medium for free to get updates from this writer.
Employer support already appears to play a substantial role. In the OECD’s 2023 Survey of Adult Skills, around 65% of adults participating in non-formal work-related learning reported receiving at least some financial support from an employer. OECD analysis also finds a positive country-level association between employer support and overall participation in adult learning. That association does not, by itself, prove causation, but it points to an important institutional reality: the burden of adaptation looks very different when workers are given time and support to learn rather than being left to absorb the cost alone.
Skills Must Be Maintained, Not Merely Acquired
The logic of continuous learning becomes easier to see in institutions where outdated competence has immediate consequences. Military reserve systems offer a useful example. A person may receive military training at 20 and remain liable for reserve service well into later adulthood, but between initial training and a later mobilisation, communications technology may change, medical procedures may evolve, new sensors may appear and entirely new capabilities may become commonplace. In such an environment, training cannot be treated as a qualification acquired once and stored indefinitely. Readiness depends on whether knowledge can still be used under present conditions.
The rapid development of unmanned systems illustrates this particularly clearly. During Estonia’s Spring Storm 2026 exercise, reservists from the 21st and 22nd Infantry Battalions received field training in the use of unmanned aerial vehicles. They practised offensive and reconnaissance operations using UAVs, as well as defence against them. Significantly, the Estonian Defence Forces noted that these reservists had already acquired basic knowledge and were receiving further training to apply those skills within their units. This is what recurrent training looks like when technology changes faster than the original training cycle: not discarding previous competence, but updating it before it becomes operationally stale.
The comparison has limits. Civilian professional education and military training operate under different incentives, obligations and consequences, and no national reserve system offers a ready-made model for the workplace. Demography, legal obligations, geography and institutional structures differ substantially. The useful comparison is narrower: competence is perishable. Skills that are not practised, refreshed or adapted to new tools cannot simply be assumed to remain current because they were once acquired or certified. In that sense, readiness is not the possession of past training; it is the maintained ability to perform under present conditions.
From Lifelong Learning to Lifelong Readiness
Switzerland institutionalises recurrent readiness through its militia system, under which military personnel return periodically for further service and training. Οn 1 March 2025, the Swiss Armed Forces counted 146,718 personnel assigned to formations. Under the militia model, soldiers are generally called up for annual three-week refresher courses, with longer periods for NCOs and officers.
Singapore offers a more direct bridge between military training and the civilian skills economy. In 2026, the Ministry of Defence and the Singapore Armed Forces announced that, in partnership with SkillsFuture Singapore, selected SAF training institutes would begin trialling a Career and Skills Passport in the second half of the year. Certificates acquired during full-time and operationally ready National Service would be recorded in the passport and mapped to industry-related skills, creating a government-verified record that could support civilian employability. The significance goes beyond the administrative recognition of military qualifications. It treats skills acquired in one institutional setting as capabilities that can be identified, translated and carried into another. Learning, in this model, does not belong exclusively to the classroom, the workplace or the military. It becomes part of a person’s accumulated capacity across different stages of adult life.
These systems are not interchangeable, and none offers a blueprint that can simply be transferred from one country or institution to another. Their value here lies in the principle they reveal: recurrent learning can be understood not merely as an educational programme, but as a way of maintaining capability over time. The wider implication extends far beyond defence. A doctor, engineer, teacher or technician cannot rely indefinitely on what was learned at 22 simply because a qualification was once awarded. Institutions face the same problem. Initial training records what someone was prepared to do at a particular moment; it does not guarantee that the same knowledge will remain sufficient decades later. The real question, then, is not whether learning continues after qualification, but whether societies treat that continuation as a private responsibility or as part of the infrastructure that keeps people and institutions capable.
The Learning Society
For generations, governments concentrated educational investment at the beginning of life. Children went to school; young adults entered universities, colleges or vocational training; the education system produced graduates and the labour market absorbed them. The model was built around a relatively clear division: education prepared people for work, while work rewarded and developed the knowledge education had already provided.
That division is becoming harder to sustain. Employees increasingly become learners, workplaces become educational environments, and employers become providers or financiers of training. Governments therefore have to plan not only for seventeen-year-olds choosing qualifications, but also for forty-five-year-olds whose occupations are changing around them. The result may be a transition from an educated society to a learning society. The distinction is more than semantic: an educated society is organised around what people have already learned; a learning society must also create the conditions in which people can continue acquiring, renewing and applying competence throughout adult life.
An educated society asks what you studied. A learning society must also ask what you are capable of learning next.
There is something potentially liberating in this idea. A person’s future need not be permanently determined by a decision made at 18 or 22. People can change profession, build new competences, return to study and acquire skills that did not even exist when they first entered employment. In a labour market shaped by technological change, the possibility of learning again can become a form of professional freedom: the ability to move rather than remain trapped inside the limits of an earlier qualification.
But freedom to adapt is meaningful only when people have a realistic capacity to exercise it. A society that requires constant adaptation has to decide whether it will provide the time, institutions and financial support that make continued learning possible, or simply transfer the responsibility to individuals and tell them to keep up. That choice matters because lifelong learning can widen opportunity, but it can also reproduce existing inequality if the people with the greatest access to time, money and employer support are the ones best able to benefit from it.
A twenty-five-year-old entering the labour market today may still be working forty years from now. Almost everything discussed in this article follows from the simple fact that the knowledge with which that career begins is unlikely to survive those forty years unchanged. Technologies will evolve, occupations will be reorganised and some skills will lose value while others emerge. The challenge is therefore larger than persuading individuals to keep learning. It is to build institutions capable of supporting people through repeated cycles of competence, disruption and renewal.
Perhaps, then, the defining skill of the twenty-first century will not simply be what we know, but whether we can learn, adapt and become competent again.
References
Cedefop. (2025). Skills empower workers in the AI revolution: First findings from Cedefop’s AI skills survey. Publications Office of the European Union.
Estonian Defence Forces. (2026, May 16). Reservists are receiving training for unmanned aerial vehicles during Spring Storm 2026.
European Commission. (n.d.). Adult learning initiatives. European Education Area.
Eurostat. (2026, September 14). SDG 4: Adult learning on the rise in the EU. European Commission.
Ministry of Defence Singapore. (2026, February 27). Updates to initiatives for NSFs and NSmen.
OECD. (2025). Trends in adult learning: New data from the 2023 Survey of Adult Skills. OECD Publishing.
Swiss Armed Forces. (2025, November 13). Effectifs de l’armée 2025 : le défi des départs anticipés.
von Felten, S. (2024). Future-oriented military leadership concepts in the Swiss Armed Forces. stratos, 2/24, 44–49.≠
