The Abundance Illusion
Why Tech Utopias Keep Appearing — and Why They Rarely Arrive
We recently came across a widely shared newsletter describing a private conversation at a Silicon Valley gathering. The claim was dramatic but familiar: within a decade, artificial intelligence and robotics will make work optional. Governments will introduce universal basic income to manage the transition. Prices will collapse as automated production floods the economy with cheap goods. Eventually we will reach a kind of post-scarcity world where people no longer need jobs at all. Nirvana is just around the corner.
The author presented the argument as if it were an insider’s glimpse into the future — the sort of thing ‘most people won’t see until it has already happened’. But the more interesting thing about the article is not its prediction, but its familiarity. We’ve heard it all before! And the suspicion is that we are hearing it again, now, because those who control the AI ‘game-board’ want us all to buy-in to the vision so that we give them a free hand vis-à-vis regulation, which means more money for them! Too cynical??
Whenever a powerful new technology arrives — steam engines, electricity, computers, the internet — a particular story tends to follow. The technology increases productivity dramatically. Human labor becomes unnecessary. Scarcity disappears. Society enters an age of abundance. And yet, somehow, that future never quite arrives.
This is not to say that technology does not transform economies. It does, often profoundly. But the relationship between technological capability and social reality is rarely as straightforward as techno-optimists assume. To understand the current wave of AI optimism, it helps to place it in a much longer historical context.
The Long History of the “End of Work”
Predictions of a work-free society are almost as old as industrial capitalism itself. When the first industrial machines began replacing human labor in the early nineteenth century, many observers assumed the end point would be a world where work was no longer necessary. If machines could produce goods more efficiently than people, then surely the logical conclusion was that machines would eventually do everything.
Some early social reformers embraced this idea enthusiastically. Industrial productivity, they argued, would generate such abundance that people would be freed from the need to work long hours simply to survive. Instead of laboring in factories, people could pursue education, creativity, or civic life.
Even Karl Marx, writing in the mid-nineteenth century, believed that industrial technology would eventually make scarcity obsolete. In his vision of a post-capitalist society, production would be so efficient that individuals would be free to pursue whatever activities they wished — hunting in the morning, fishing in the afternoon, and criticizing after dinner.
“The realm of freedom actually begins only where labour which is determined by necessity and mundane considerations ceases.” Karl Marx. Capital, Vol. III, Chapter 48
The prediction was wrong in its details but not entirely in its intuition. Industrialization did indeed generate enormous productivity gains. Over the next century and a half, living standards rose dramatically. But work did not disappear. Instead, work changed. New industries emerged. Consumption expanded. The economy reorganized itself around new forms of labor. This pattern repeated itself throughout the twentieth century.
Keynes and the Fifteen-Hour Week
One of the most famous predictions of a leisure society came from the economist John Maynard Keynes. In 1930, Keynes wrote an essay titled Economic Possibilities for our Grandchildren. In it he argued that productivity growth would eventually allow societies to meet their material needs with far less labor. Within a hundred years, he predicted, the average working week might fall to around fifteen hours.
“For the first time since his creation man will be faced with his real, his permanent problem — how to use his freedom from pressing economic cares.” John Maynard Keynes.
— Economic Possibilities for our Grandchildren (1930)
Keynes was not naïve. He understood that productivity increases do not automatically translate into leisure. But he believed that once basic needs were satisfied, societies would gradually choose more free time over more consumption. The productivity part of his prediction turned out to be largely correct. Since 1930, productivity in advanced economies has increased by several multiples. But the working week never fell to fifteen hours.
Instead, economies became more productive, societies expanded their expectations of what constitutes a normal standard of living. Larger homes, more travel, better healthcare, more education, more entertainment, more services. Productivity gains did not eliminate work, they enabled a much richer form of economic life.
The Automation of the 1960s, 1970s, 1980s and beyond
The debate resurfaced again in the 1960s with the rise of computers and early automation. Policy makers in the United States began worrying about the prospect of widespread technological unemployment. If computers could process information faster than humans, and machines could increasingly manage production, perhaps large parts of the workforce would become permanently unnecessary.
Government reports from the period warned of a future where automation might produce mass unemployment. At the same time, others predicted a leisure society. Technology would take care of routine tasks while humans pursued culture, creativity, and self-development. But the same thing happened again.
Computers transformed the economy. Entire industries were reshaped or replaced. Yet employment did not disappear. Instead, new sectors emerged — software, digital services, financial technology, media, information industries — many of which were unimaginable only decades earlier.
The pattern repeated once more during the rise of the internet in the 1990s. Digital technology was supposed to create frictionless markets and limitless information. Some commentators argued that scarcity itself was ending. What actually emerged was something more complex: a digital economy with immense wealth creation, but also new forms of inequality, concentration of power, and economic disruption.
And now the argument has returned again — this time attached to artificial intelligence.
The Current Wave of AI Abundance
Today’s techno-optimist narrative follows a familiar structure. Artificial intelligence will automate cognitive work just as machines automated physical labor. Software will write software. AI agents will manage operations. Robots will manufacture goods. Production costs will collapse.
As automation spreads, prices fall. Eventually goods and services become so cheap that even a modest basic income provides a comfortable life. Work becomes optional. The story is appealing for obvious reasons. It promises a future where technological progress solves one of the oldest problems in economics: scarcity. The embedded assumption is that technological capability automatically translates into social outcomes. History suggests that it does not.
“My prediction is that work will be optional. It’ll be like playing sports or a video game or something like that. If you want to work, [it’s] the same way you can go to the store and just buy some vegetables, or you can grow vegetables in your backyard. It’s much harder to grow vegetables in your backyard, and some people still do it because they like growing vegetables.” Elon Musk
The Missing Variable: Institutions
Technology determines what becomes possible. Institutions determine what actually happens. This distinction matters because productivity gains do not automatically distribute themselves across society. When technology increases productivity, the benefits tend to flow initially to those who control the technology … the firms, investors, and organizations that deploy it.
Only later, often through political negotiation or institutional change, do those gains spread more widely. The Industrial Revolution dramatically increased productivity, but it took decades of labor movements, regulatory reforms, and social policies before the benefits were widely shared.
The same dynamic appeared in the twentieth century. The rise of mass prosperity in many Western countries after the Second World War was not simply the result of technological progress. It was also the result of institutions — welfare states, public education systems, labor protections, and progressive taxation. Without those structures, productivity gains might have produced a very different social outcome.
This is why the AI abundance narrative often feels incomplete. It focuses on technological capability but largely ignores institutional dynamics. Yet those dynamics will almost certainly determine the shape of the future.
Conditions for a Post-Scarcity Society
For a genuine post-scarcity economy to emerge — one where most people no longer need to work — several conditions would have to be met simultaneously. First, productivity would have to increase dramatically across large parts of the economy. AI and robotics would need to automate not just narrow tasks but broad systems of production. Second, the cost of producing goods and services would have to fall substantially. This is plausible in some areas, particularly digital services and certain forms of manufacturing. But many major costs in modern economies are not purely technological. Housing, infrastructure, healthcare systems, energy grids, and regulatory frameworks all shape economic costs in ways that automation alone cannot easily eliminate. And thirdly, the gains from automation would need to be distributed widely enough that most people could maintain a decent standard of living without traditional employment.
This is where the real challenge lies. In a market economy, firms typically operate with the objective of maximizing returns to shareholders. If automation dramatically increases productivity, the immediate effect is likely to be higher corporate profits and lower demand for labor.
Then unemployment becomes less a lifestyle choice and more an existential crisis.
Without mechanisms for redistribution, the result would not be post-scarcity abundance but increasing concentration of wealth. Some techno-optimists assume that governments will simply introduce universal basic income to solve this problem. But redistribution at that scale raises complex political and economic questions. Governments would need to fund those payments through taxation or public ownership of productive assets. Companies might resist or relocate if the tax burden becomes too heavy, hollowing out that country’s tax base just when it needed it the most.
None of these challenges are insurmountable. But they are political questions, not technological ones. And political transitions rarely happen quickly.
Beyond institutional dynamics, several other barriers make the rapid arrival of a post-work society unlikely. One is the persistence of scarcity in physical systems. Even if information and software become extremely cheap, the physical world still operates under constraints — land, energy, materials, infrastructure etc. (as has been so clearly demonstrated by Washington’s ‘war of choice’ in the Middle East). Another barrier is demand itself. As productivity increases, societies often expand their expectations of what constitutes a good life. New goods and services emerge, creating new forms of work. A third barrier is cultural. Work is not simply an economic activity. It also provides identity, structure, social interaction, and status. Even in wealthy societies where basic needs are largely met, people continue to work — not only for income but for meaning.
This does not mean work must always take the same form. But it does suggest that a world where most people simply stop working altogether may be less appealing in practice than it appears in theory.
A More Plausible Future
If we step back from the extremes — neither techno-utopia nor technological collapse — a more plausible future begins to emerge. Artificial intelligence is likely to transform knowledge work in ways comparable to how earlier technologies transformed physical labor. Many tasks will become heavily automated. Productivity will increase. But rather than eliminating work entirely, AI will probably change the nature of work.
Many roles may shift toward supervision, coordination, interpretation, and creativity — activities that involve working with AI systems rather than competing directly with them. But there may not be scope to keep everyone gainfully employed in supervisory roles.
“There will be some things we reserve for ourselves. But in terms of making things and moving things and growing food, over time those will be basically solved problems” Bill Gates
Productivity gains could allow societies to reduce working hours gradually. Shorter workweeks, longer education periods, and more flexible career paths may become common. But will workers work less for the same pay … or does less work equal less income?
Governments may experiment with new forms of social protection — perhaps partial basic income schemes, expanded public services, or collective ownership of certain technological infrastructures. Firms themselves may evolve as well. Some organizations may adopt broader stakeholder models, employee ownership structures, or cooperative governance.
None of this looks like the dramatic “end of work” promised by techno-optimists. But it may still represent a significant shift in how economies function.
Interestingly, the techno-optimists may be right about one thing: if automation continues to expand, the real challenge may not be economic but existential. Modern societies have organized much of life around work. Work structures time, provides identity, and shapes social status. If technology gradually reduces the centrality of work in economic life, societies will need new ways to organize meaning and purpose. This is not a trivial challenge. In fact, it may turn out to be one of the defining cultural questions of the twenty-first century.
Why the Utopian Narrative Persists
So why do predictions of post-scarcity abundance keep appearing?
Partly because technological progress really does expand the boundaries of what is possible. Each wave of innovation genuinely transforms the economy in ways that would have seemed extraordinary only decades earlier. But there is also a deeper psychological appeal. The idea that technology might eventually free humanity from the necessity of labor has a powerful resonance. It echoes the philosophical dreams of a society where survival is no longer the central struggle of life. Each new technology briefly makes that dream seem close again. Any the cynic might suggest that it is held out like a carrot in the front of the donkey … the promise of a future utopia if we let the masters of the AI universe do their thing.
Yet history repeatedly reminds us that technology alone does not determine the future. The future is shaped by the interaction between technology, institutions, culture, and politics. Artificial intelligence will almost certainly transform the economy in profound ways. But the path from technological capability to social outcome is rarely straightforward.
The abundance imagined by techno-optimists is not impossible. It is simply not inevitable (or even likely). And whether it emerges will depend far less on the capabilities of our machines than on the choices we make about how to organize the societies in which those machines operate.




Adam and David, this is a much-needed corrective to the techno-optimist narrative. The distinction between technological capability and institutional reality is the piece that most abundance discussions willfully ignore.
I'd add a practical dimension from the infrastructure side that reinforces your argument about physical-world constraints.
I work in IoT and edge AI deployment -- the layer that actually connects AI to physical production systems. And the gap between what AI can do in a demo and what it can do at scale in the real world is precisely the kind of barrier you're describing, except it's technical rather than institutional.
The abundance thesis assumes AI and robotics will collapse production costs across housing, food, energy, and manufacturing simultaneously. But each of those sectors has deep physical infrastructure dependencies that don't respond to software-like scaling. A factory deploying AI-powered predictive maintenance still needs sensors on every critical asset, reliable connectivity across the shop floor, edge compute for real-time inference, and integration with legacy control systems that are often 15-20 years old. That buildout costs real money, takes real time, and encounters real-world friction that no foundation model can shortcut.
Your Keynes parallel is particularly apt. He was right about productivity but wrong about how societies would allocate the gains. I think the same pattern is playing out with AI: the productivity gains from automation are real but will flow disproportionately to capital owners and the companies that control the deployment infrastructure, not to workers or consumers. The cost of deploying physical AI at scale -- connectivity, hardware, integration, maintenance -- creates a capital barrier that naturally concentrates benefits.
The institutional question you raise is the right one: who designs the redistribution mechanisms? But I'd add that there's also a timing question that's underappreciated. The physical infrastructure buildout required for AI to actually deflate real-world costs (not just digital services) will take 10-15 years minimum. During that gap, we get the worst of both worlds: job displacement from AI adoption in white-collar work, without the cost deflation in physical goods that's supposed to make it bearable.
That valley between displacement and deflation is where the political risk concentrates. And it's a much longer valley than the techno-optimists admit.