A labour market in transition, not in retreat
For several years, debates about the future of work have oscillated between two simplistic claims: that offices are finished, or that remote work was a temporary aberration soon to be corrected by managerial insistence. Neither view survives contact with the evidence. What is emerging instead is a more complex reorganisation of work, one in which location, technology and organisational design are being renegotiated at the same time.
The shift matters because it reaches beyond office etiquette. It touches productivity growth, wage formation, urban transport, commercial property, family life and the bargaining power of labour. Research from the International Labour Organization has shown that home-based and remote work expanded sharply during the pandemic but remained highly unequal by occupation, gender and country income level. Meanwhile, data gathered by academic labour economists suggest that hybrid arrangements have endured because they can preserve much of the co-ordination value of in-person work while offering workers more control over time.
The result is not a frictionless digital labour market. It is a more segmented one. High-autonomy, screen-based occupations have gained flexibility. Place-dependent work in health, logistics, manufacturing, retail and care has not. The future of work therefore cannot be analysed merely as a story of freedom. It is also a story about who can convert flexibility into leverage, and who cannot.
“The future of work will be decided less by where people sit than by whether institutions can turn flexibility into productivity and security at the same time.”
Hybrid work is settling into the mainstream
The strongest evidence now suggests that hybrid work, not fully remote work, is the durable organisational form for many professional occupations. Studies published by the National Bureau of Economic Research and survey work associated with the Stanford-led research on working arrangements show that workers value flexibility highly, often close to the value of a meaningful pay increase. Employers, for their part, have become more selective. Many have concluded that fully distributed work can weaken mentoring, innovation spillovers and social cohesion, but that rigid attendance rules impose their own costs in attrition and recruitment.
This helps explain the apparent paradox in office life. Workers are returning, but not reverting. Occupancy data in major business districts remain below pre-pandemic norms in many countries, even where hiring has stabilised. Days in the office are increasingly clustered around collaboration, training and client interaction, while concentrated individual work is more likely to take place elsewhere. The office is becoming a more specialised asset: less a default container for all tasks than a site for a subset of high-value interactions.
This reallocation carries strategic implications. Firms that benefit from hybrid work are not merely permitting flexibility; they are redesigning workflows around it. That means clearer documentation, more explicit performance metrics, stronger management practices and investment in digital infrastructure. Hybrid work succeeds where ambiguity falls and trust rises. It fails where it becomes a disguised form of presenteeism, with workers expected to be permanently available regardless of location.
Artificial intelligence will change tasks before it destroys jobs
The loudest claims about automation often focus on headline job losses. The more plausible near-term effect is subtler but still profound: the recomposition of tasks within existing jobs. The Organisation for Economic Co-operation and Development has argued that generative artificial intelligence is likely to affect a broad range of cognitive occupations, especially those involving language, pattern recognition and standardised decision support. But exposure is not destiny. In most sectors, technologies are adopted unevenly, constrained by regulation, data quality, workflow integration and the simple difficulty of changing how organisations operate.
The future of work will be decided less by where people sit than by whether institutions can turn flexibility into productivity and security at the same time.
Recent work from institutions such as the International Monetary Fund and McKinsey Global Institute points in the same direction. A significant share of current work activities can be automated or augmented, yet the net employment effect depends on demand growth, complementary investment and reskilling. In practice, automation often removes routine components while increasing the value of judgment, client management, systems thinking and domain expertise.
This matters for white-collar work in particular. Many professionals will not be replaced outright; they will be asked to supervise, verify and orchestrate machine outputs. That could raise productivity, but it may also alter career ladders. Entry-level roles traditionally used for apprenticeship may shrink if first drafts, basic coding, document review or standard analysis can be generated automatically. Firms and policymakers should therefore pay attention not only to displaced workers, but also to missing pathways into skilled work.
“Artificial intelligence is likely to transform jobs first by unbundling tasks, not by eliminating occupations in one dramatic sweep.”
Productivity remains the unresolved question
The future of work will ultimately be judged by productivity, because productivity is what allows wages to rise without reigniting inflation. Yet productivity outcomes from both hybrid work and digital tools remain mixed. The pandemic accelerated adoption of collaboration software, cloud systems and process automation, but aggregate productivity growth in many advanced economies has remained disappointing. This is not necessarily evidence of failure. Large technologies often produce gains only after organisational adaptation, complementary investment and the exit of outdated practices.
The historical lesson is that diffusion lags invention. The mere presence of a powerful tool rarely guarantees economy-wide gains. Managers must rethink processes, workers must build new habits, and firms must absorb the costs of transition before benefits appear in national accounts. Research by the OECD and the World Bank has long suggested that productivity gaps between frontier firms and laggards owe as much to management quality and diffusion capacity as to invention itself.
That is why the most important workplace technology may not be the most glamorous one. Better knowledge management, cleaner data, interoperable systems and more disciplined project management can produce larger gains than splashier experiments. The future of work is not simply about adding intelligence to software. It is about reducing friction in institutions that are often still surprisingly analogue.
Cities are adapting rather than fading
Predictions that remote work would permanently hollow out cities were always overstated. Big urban areas retain powerful advantages: dense labour markets, specialised services, research institutions and the chance encounters that support entrepreneurship and learning. What has changed is the rhythm of demand. Transport peaks are flatter. Retail and hospitality income in central business districts is more uneven. Office demand is shifting towards higher-quality, better-located space while older stock struggles.
This has important consequences for urban policy. Cities that prosper in the next decade will not merely wait for commuters to return five days a week. They will repurpose underused commercial space, improve transport flexibility, expand housing where demand remains strong and make central districts attractive for more than office work. The geography of work is becoming more distributed, but distribution does not mean dispersion into irrelevance. It means that successful cities must become more multifunctional.
There is also a social dimension. If affluent professionals work remotely from desirable neighbourhoods while lower-paid workers still bear the cost of commuting, inequality can become more spatially entrenched. Access to affordable housing, childcare and digital infrastructure will therefore shape who benefits from new work patterns. A city-friendly future of work requires more than broadband and coffee shops; it requires practical inclusion.
The labour market is tightening structurally in some places
Artificial intelligence is likely to transform jobs first by unbundling tasks, not by eliminating occupations in one dramatic sweep.
Another reason the future of work cannot be reduced to technology is demography. In many advanced economies, workforce growth is slowing as populations age. The IMF, OECD and World Bank have all pointed to ageing as a drag on labour supply and potential output. This strengthens the bargaining position of some workers, particularly in sectors facing persistent shortages, from healthcare to engineering to skilled trades. It also increases the urgency of raising participation among groups that remain underrepresented or constrained by care responsibilities, disability or geographic mismatch.
Demography changes employers’ incentives. Where labour becomes scarcer, retention matters more. Flexible schedules, better management and investment in training become less a matter of generosity than of necessity. Yet ageing also sharpens fiscal pressures. If social spending rises while productivity disappoints, governments may find it harder to finance the very systems of retraining and social protection needed for a smoother transition.
This is one reason immigration policy remains central to labour-market strategy, even when politically contentious. Economies with weak domestic labour-force growth will struggle to sustain services, innovation and tax revenues without some combination of migration, higher participation and faster productivity. The future of work, in other words, is partly a question of who is able to work, not merely how they work.
Skills policy must move from credentials to capability
If technology is reshaping tasks, then education systems built around one-off front-loaded credentialing are poorly matched to the problem. Workers will need recurrent skill renewal across longer careers. Yet many public training systems still struggle to identify which skills are in demand, to reach mid-career workers, or to verify shorter forms of learning in ways employers trust.
The World Economic Forum’s labour-market analysis and OECD skills work both suggest that analytical, social and adaptive capabilities are becoming more valuable alongside technical literacy. This does not mean everyone must become an engineer. It means more occupations will require a baseline ability to work with data, digital systems and automated tools, while human skills such as persuasion, care, negotiation and leadership become more economically visible.
Policy should therefore focus less on fashionable lists of “jobs of the future” and more on transferable capability. Apprenticeships, modular learning, mid-career subsidies and better labour-market information can all help. So can stronger links between employers and educators. The critical test is whether workers can move across adjacent roles as tasks change. Resilience in the labour market depends on permeability, not prophecy.
“The central skills challenge is not predicting a fixed set of future jobs, but building systems that let workers keep moving as tasks evolve.”
Management is becoming a technology in its own right
One underappreciated consequence of hybrid work and automation is the rising value of competent management. In more distributed organisations, managers must communicate goals more clearly, evaluate outcomes more fairly and maintain cohesion without relying on informal visibility. They also need enough technical literacy to understand how automated systems are affecting workflows, risks and worker morale.
This sounds mundane, but it is economically significant. A long body of research has linked management quality to productivity differences across firms and countries. Poorly managed organisations often fail not because workers resist change, but because leaders introduce tools without redesigning incentives, training or accountability. The result is frustration masquerading as innovation.
As routine supervision is partly digitised, the manager’s role shifts from monitoring attendance to creating clarity, trust and development. That requires emotional discipline as much as operational skill. It also suggests that future labour disputes may concern algorithmic fairness, surveillance and performance data as much as pay and hours. The politics of management will become more technical, and the technology of management more political.
The central skills challenge is not predicting a fixed set of future jobs, but building systems that let workers keep moving as tasks evolve.
Social protection was built for a different era
Even before the latest wave of automation anxiety, many labour markets were already characterised by fragmented careers, contract work and weaker attachment to a single employer. Hybrid work does not create that reality, but it can intensify it by widening cross-border contracting and making labour markets more contestable. At the same time, platform-mediated work and self-employment continue to expose gaps in systems built around stable payroll employment.
This raises an old but unresolved question: how portable should benefits be? Health insurance, pensions, unemployment support and training rights often remain tied to employment categories that no longer capture how many people actually work. The ILO has repeatedly argued for extending social protection coverage and adapting labour standards to new forms of employment. That is not simply a welfare concern. It affects labour mobility, entrepreneurship and the willingness of workers to embrace technological change.
Countries will differ in institutional design, but the direction is clear. Where risk is individualised too aggressively, workers become more defensive and transitions harder. Where protection is portable and predictable, economies can adjust more quickly. The future of work will be shaped not just by frontier firms, but by the administrative competence of states.
The office debate hides a broader question of power
Arguments over return-to-office policies often appear cultural, even generational. Beneath them lies a more basic issue: who has discretion over time, monitoring and performance. Flexibility can be empowering, but it can also become a mechanism for cost-shifting, with workers absorbing expenses for space, energy and equipment while remaining permanently reachable. Conversely, in-person work can support solidarity and mentorship, but it can also reinforce hierarchy through visibility and exclusion.
The distribution of power will therefore depend on governance. Are performance metrics transparent? Are workers consulted on technological deployment? Do managers use surveillance tools proportionately? Are promotion pathways available to those who are not physically present every day? These questions matter because organisational norms, once established, can persist long after the original technological shock has faded.
There is no single optimal settlement. Different sectors require different balances of autonomy and co-ordination. But workers are likely to accept change more readily when they can see the rationale, share in the gains and trust the rules. The future of work is not a destination delivered by software; it is a negotiated order.
What a durable settlement might look like
The most credible future of work is neither utopian nor reactionary. It does not assume that digital tools will liberate everyone from drudgery, nor that prosperity depends on recreating the habits of 2019. A durable settlement would combine several elements: hybrid arrangements where they improve performance, strong in-person institutions where work depends on them, investment in complementary skills, and social protections that make labour mobility less perilous.
For firms, that means choosing technology with a clear view of workflow and workforce effects rather than adopting it as theatre. For cities, it means redesigning central districts for mixed use and reducing the penalties of mobility. For governments, it means treating training, childcare, migration and social insurance as labour-market infrastructure rather than peripheral spending. For workers, it means accepting that careers are likely to become more fluid, while insisting that fluidity need not mean precarity.
The old settlement of work was built around a stable workplace, linear progression and benefits tied to long-term employment. The new one will be looser, more digital and more geographically flexible. Whether it is also more productive and more fair will depend on choices that are institutional, not inevitable. The future of work is being written in code and concrete alike. The code matters, but the concrete — offices, housing, transport, schools and public systems — will decide who can actually live with the result.
