Beyond the End of Work: Adapting to the New Economic Reality

I have already written about how some future version of AI will lead to the end of cognitive and then physical labour being economically viable. But, then what? How does a post-labour economy operate? Who works? Who, or what, produces products and services? Where is wealth accumulated? Who, or what, pays taxes? How do citizens receive income to buy good and services? 

Queue for UBI

These are just a few of the questions covered below.

The Decoupling of Labour and Income

At the turn of the millennium, wages still accounted for nearly 58% of output in OECD economies. By 2024, this figure had dropped below 54%—and that decline occurred before generative AI had truly arrived. In the media there are a growing number of stories about AI replacing workers, or large companies developing strategies to limit the recruitment of human workers. There are fluctuations in strategy (e.g. Klarna and Duolingo), but the direction of traffic is clear. 

Today, 42% of global working hours are susceptible to automation, according to the Oxford-OECD index, a dramatic rise from 31% just five years ago. MIT economist David Autor highlights a troubling pattern: each percentage-point drop in labour share of GDP reduces consumer demand by 0.6% within two years, squeezing household finances and eroding economic resilience.

The critical metric to watch is labour share of GDP. At the current trajectory, by the late 2020s wages will no longer represent the majority of economic activity—a shift that fundamentally undermines the wage-based social contract.

Roadmap to a Post-Labour Economy (2025–2050)

2025–2027: The Age of Cognitive Automation:  Generative AI is now mainstream and it may just be a technology stepping stone on a path to far more powerful AI technology that will lead to artificial general intelligence (AGI) and then artificial superintelligence (ASI). 

Roadmap to the end of work

By 2027, Goldman Sachs anticipates AI will significantly reshape 300 million jobs globally. This first wave targets cognitive work—marketing, accounting, law, programming—with productivity gains rapidly turning into displacement.

2028–2032: Robotics Reaches Its “ChatGPT Moment”: Physical automation trails cognitive tasks, but this is set to accelerate. China’s robot installations surged to 290,000 in 2024 alone and are projected to double by 2028, exceeding Japan’s industrial robot density for the first time. Driven by severe demographic pressures, China's national strategy targets 470 robots per 10,000 workers by 2027—a critical tipping point beyond which human labour becomes optional in manufacturing. 

Compare a robot costing, say, £25,000 to car finance and the cost is ~£500 pm for a worker that can operate 24x7 depending on how it is charged, and maintenance. 

2033–2037: Demand Shock and Policy Inflection: By 2035, according to McKinsey's forecasts, over half of routine cognitive roles and a third of physical jobs will vanish, severely impacting consumer spending. Ironically, this will be very much the case at McKinsey as management consulting has no regulatory protection. Governments will be confronted with an unprecedented dilemma: introduce permanent income support measures, redistribute asset ownership, or risk prolonged recession or even social unrest.

2038–2050: S-Curve Maturity: At full maturity, around mid-century, there will be roughly one billion robots worldwide—approximately one for every ten humans—and global AI-service revenue will reach trillions annually. Human labour becomes largely redundant as energy, data, and intellectual property—rather than workforce numbers—define economic competitiveness. Countries failing to create new income distribution frameworks face persistent low demand, economic stagnation, and heightened social fragility.

The New Sources of Wealth: AI, Robotics, and Intangibles

The wealth generated in a post-labour economy will primarily stem from intangible assets: GPU compute hours, proprietary data, and algorithmic intellectual property. By 2030, global spending on AI infrastructure will surpass $4 trillion, according to Accenture, driven by data centres and cloud computing. Margins are staggering; services like GPT-4 achieve over 80% profitability per API interaction.

Demographic challenges, especially in China, Korea, and Japan, accelerate robotic adoption. Foxconn’s main iPhone factory in Zhengzhou reduced its human workforce by 17% in a single year while simultaneously increasing output—clear proof that wealth accrual now favours the owners of IP, robots, and data infrastructure.

Why the Economic Multiplier Matters More Than Ever

In today’s economy, every pound earned by a worker doesn’t just benefit that individual—it cascades through the domestic and other economies. Workers typically spend most of their income on goods and services, creating a ripple effect where one person’s spending becomes another’s wages. Economists call this the ‘economic multiplier,’ and it’s the engine behind thriving local economies.

But as automation reduces wages, profits increasingly accumulate with businesses and wealthy investors, who tend to spend a much smaller fraction of their extra income. Without action, this shift slows the multiplier effect, risking stagnation in local economies.

How can we keep economic activity flowing in a world with less wage income? By actively distributing the gains from AI and robotics through direct cash payments and shared dividends. For example, even modest universal payments of a few hundred pounds per month significantly boost consumer spending, as most households spend over 80% of additional income. Similarly, shared dividends—like those seen in Alaska, where oil revenues go directly to residents—quickly flow back into the economy.

Ultimately, to maintain a vibrant economy, we must ensure profits from technology-driven productivity reach ordinary people who spend rather than stockpile wealth. In a post-labour world, carefully engineered income and ownership distribution isn’t merely fair—it’s vital to sustaining economic momentum.

Models for Distributing the AI Dividend: UBI, UHI, and Asset Ownership

Universal Basic Income (UBI): UBI is straightforward but politically challenging. Modelling by the Roosevelt Institute suggests a modest monthly UBI of $300 would significantly reduce poverty in the US at an annual cost of about 3% of GDP—a practical yet powerful tool.

Finland's UBI trial (2017–2018) provided 2,000 unemployed citizens with €560 monthly, resulting in improved mental well-being, reduced stress, and a modest increase in employment, highlighting UBI’s potential to enhance life quality and stability in the face of economic uncertainty.

Universal High Income (UHI): Sounds great—don't work, but receive a high income! Mustafa Suleyman proposes a bold alternative to traditional UBI, known as Universal High Income (UHI). Rather than relying on general taxation or government spending, UHI would be directly funded by licensing fees charged to companies developing and operating "frontier" AI models—such as GPT-4 or Google's Gemini. Suleyman estimates these fees could collectively generate up to $1.5 trillion annually by 2035. At scale, this revenue could translate into approximately $10,000 per adult per year across the G7 economies, offering a meaningful economic floor that far exceeds most current basic income proposals. 

However, achieving this vision hinges on securing unprecedented global consensus and creating effective international enforcement frameworks, as companies could otherwise relocate to jurisdictions without such licensing requirements. Given the current trade tariff battles, and even wars, this will be a big ask. However, if successfully implemented, this approach could establish a robust link between AI-driven prosperity and broad societal benefit, transforming AI from a potential driver of inequality into a cornerstone of shared economic security.

Dividend Trust Models: The author Daniel Susskind advocates establishing sovereign AI funds financed by modest levies on AI-related revenues. For instance, a mere 2% charge on cloud-compute transactions could yield $250 billion annually by 2030, enough to seed meaningful dividends.

Regardless of model, the fundamental solution is clear: transition from wages to asset-based income. The pivotal metric is the proportion of household income derived from dividends rather than wages. When dividends surpass a third of household income, society will have successfully transitioned to a sustainable post-labour economy.

Maintaining Aggregate Demand in Advanced Economies

Harvard economist Jason Furman warns of severe demand-driven recessions in economies reliant on consumption (70% of US GDP). With real median wages nearly flat since 2019 despite booming corporate profits, advanced economies risk a consumption collapse without intervention.

Emerging policy innovations include:

• AI-dividend Royalties: Canada proposes micro-royalties on commercial AI usage funnelled into local trusts.

• Share-Grant Credits: New Zealand’s recent legislation allowing corporate tax payments via equity contributions to public funds, creating shared prosperity through public ownership.

• Economic Agency Dashboards: Real-time county-level tracking of income sources—wages, transfers, dividends—to inform targeted policy interventions.

Watch two key metrics: median disposable income and median wage levels. Sustained divergence, with disposable income rising despite falling wages, signals successful economic restructuring.

The Impact on Low-Cost Labour Economies

Globalisation’s traditional reliance on cheap human labour, both physical and cognitive, is rapidly eroding due to breakthroughs in robotics and generative AI. Robotic operating costs already sit below minimum wage levels in many emerging economies, and automation is poised to eliminate millions of manufacturing jobs—40 million in textiles alone within the decade, according to the World Bank. 

End of the knowledge work cost advantage: However, the previously unassailable advantage held by countries in affordable knowledge-based services, from call centres to software development, will diminish under the advancing capabilities of generative AI and future, more advanced AI. In a world with AGI, there will no zero economic advantage. Regions like India, Vietnam, and the Philippines, which currently host millions of outsourced knowledge workers, are highly exposed as AI-driven platforms deliver comparable quality at a fraction of the cost.

Despite this challenge, developing economies retain strategic opportunities to reposition themselves:

• Robot-Leasing Consortia: By financing robots and AI systems centrally and leasing them to manufacturers and service providers, countries can secure recurring income streams for national development trusts, offsetting lost tax revenue from declining employment.

• Equity-for-Market Access Deals: Countries can leverage their large consumer bases and market access rights by negotiating equity stakes or revenue shares in foreign companies seeking to deploy automation, as Gujarat, India, successfully did with German automakers in 2024.

• Vocational Transformation: Countries like Vietnam and the Philippines are beginning to repurpose traditional vocational training programs towards high-value skills such as robotics maintenance, AI oversight, and algorithmic governance, preserving employment relevance even as routine tasks are automated.

Crucial metrics to watch include foreign direct investment directed specifically towards automation infrastructure, and youth unemployment trends. A simultaneous rise in both indicators signals that traditional employment structures are deteriorating faster than new roles are emerging, highlighting the urgent need for proactive, equity-focused strategies to ensure the benefits of automation flow broadly rather than concentrating narrowly.

Society After Work: Redefining Identity, Well-being, and Purpose

Work traditionally offers identity and purpose alongside income. Its erosion introduces a critical social challenge. Susskind calls this the “meaning paradox”: more leisure time, but with less collective purpose.

Proven strategies already exist:

• Civic Contribution Credits: Barcelona successfully encourages purposeful community engagement by rewarding social contributions with convertible dividend credits.

• Adaptive Urban Spaces: Seoul transforms office space into civic, productive hubs, creating new societal roles and municipal revenues.

• Time-Sufficiency Index: Denmark tracks how citizens allocate time to personally meaningful pursuits as a measure of national well-being.

Metrics to track closely: average weekly working hours and national life-satisfaction indices. A positive correlation—fewer hours worked, increased satisfaction—reflects successful societal adaptation.

A Practical Transition Playbook

Governments:

• Test and establish a modest UBI linked to unemployment rates exceeding 10%. Fight pushback as this is not a new benefit and today's highly paid knowledge workers may soon depend on it.

• Implement dividend laws similar to Norway’s oil fund, funded through AI and robotics licensing fees.

• Publish economic-agency dashboards to provide transparent, accessible dashboards showing clearly where household incomes come from—wages, assets, or transfers—to guide better policy decisions.

Businesses:

• Voluntarily allocate equity stakes annually into local trusts for tax relief and public goodwill.

• Invest savings from automation directly into community dividends and retraining funds.

• Transform customer loyalty programmes into equity partnerships, offering patrons dividend-paying shares.

Financial Institutions:

• Develop dividend-based financial instruments and lending practices that treat dividend income as stable cash flow.

• Package robot-leasing bonds for developing markets with OECD pension fund guarantees.

Households:

• Diversify income streams across multiple dividend sources: trusts, cooperative shares, royalty rights, and AI dividends.

• Negotiate compensation packages emphasizing equity participation over wages.

Key metrics for real-time monitoring: robot density per 10,000 workers, labour’s GDP share, and household consumption rates. Together, these indicators reveal the speed and direction of transition.

Conclusion: Navigating the End of Work

We are witnessing the evolution of an economy defined not by human labour, but by physical and thinking machines that will transform our relationship with work itself. The choices we make now will dictate whether this transformation leads to prosperity or deepens inequality. Those choices will also impact societal anxiety and the likelihood of confusion and even unrest. The blueprint already exists: universal income schemes to anchor demand, and innovative asset-ownership models to democratise wealth generated by AI and robotics.

The tools are available, and the path is clear. The greatest risk isn’t technical—it’s cultural denial. Governments, organisations, and citizens must stop pretending this future is fiction. 

Acting proactively ensures the gains from technology benefit all, creating societies where stability does not depend on employment alone. The alternative—delay and denial—risks a fractured future marked by economic exclusion.

Ultimately, the end of human work need not signal crisis, but opportunity. Did we evolve to be defined by 'work'? By embracing change decisively, we can redefine progress and emerge stronger, fairer, and more prosperous. 

The real question is not whether we can afford to act, but whether we dare not to.

Thanks for reading.