Recent remarks by Elon Musk have reignited a familiar but increasingly consequential debate about artificial intelligence and the economic future. In a widely circulated clip, Musk suggests that a world of “universal high income” is not only possible, but non-inflationary — even if funded by direct government payments — because AI-driven production will expand so rapidly that it outpaces the growth of the money supply, causing deflation.

In that formulation, prices are reduced to a simple ratio: if goods and services grow faster than money, then more money can be distributed without distorting the system. Extend the logic further, and saving for retirement becomes unnecessary, work becomes voluntary, and scarcity itself begins to fade into irrelevance.

It is a striking vision — optimistic, internally coherent at a glance, and appealing in its promise to dissolve tradeoffs. It also arrives at a moment when Musk’s commercial and financial interests are increasingly tied to the scale and perceived inevitability of the AI transformation. If he becomes the world’s first trillionaire — a milestone that now appears plausible within the year — it will not merely reflect entrepreneurial success, but the extraordinary forward-looking valuations markets are placing on AI-related enterprises.

His accomplishments in engineering, manufacturing, and capital formation are, by any measure, remarkable. But technological brilliance and commercial success do not automatically confer authority in economic reasoning, particularly on questions as old and as deeply studied as money, prices, and coordination.

The difficulty with Musk’s argument begins with a deceptively simple point: issuing money does not create income in any real sense. It redistributes claims on output. Even in a world of rapidly expanding production, goods and services must still be produced, and the distribution of purchasing power (among other factors) determines who can access them.

The path by which new money enters the system is therefore crucial.

It does not arrive evenly or instantaneously. It flows through specific channels — government transfers, financial markets, credit systems — and those entry points shape the sequence in which prices adjust. Some sectors see rising demand and expanding activity first; others face higher input costs or shifting consumption patterns. These relative price changes are not incidental. They are the mechanism by which the economy coordinates production across time and space.

This is why the notion that inflation or deflation can be understood as a simple ratio of total output to total money is misleading. Prices are not set in the aggregate. They are relative, reflecting localized conditions of supply, demand, expectations, and timing.

An AI-driven expansion in output will not occur uniformly across all sectors. It may sharply reduce the cost of digital services, standardized manufacturing, or logistics optimization. But it will not make land in Tokyo or Manhattan more abundant. It will not eliminate constraints on energy infrastructure overnight. It will not expand human attention, physical presence, or positional goods. What emerges instead is not the disappearance of scarcity, but its reconfiguration — deflation in some areas, persistent or even rising prices in others.

Once that is understood, the monetary claim becomes more fragile. Injecting new purchasing power into such an uneven landscape does not simply “scale up” consumption. It redirects it. And because those adjustments occur sequentially rather than simultaneously, they can distort the signals that guide investment and production. When credit conditions are shaped by sustained monetary expansion, borrowing costs may be suppressed below levels consistent with underlying resource availability or time preferences. Projects that appear viable under those conditions may prove unsustainable once financing tightens or expectations shift. The result is not a frictionless transition to abundance, but the familiar pattern of overextension, misallocation, and eventual correction.

The idea that technological progress might eliminate the need for saving — or render retirement planning obsolete — rests on a similar misunderstanding. Even if AI dramatically reduces the cost of many goods, it does not eliminate uncertainty, time, or the need to allocate resources across the life cycle. Individuals still face unknown future conditions, changing preferences, and risks that cannot be forecast with precision.

Saving is not merely a hedge against high prices; it is a way of transferring resources across time in the face of uncertainty. That function does not disappear simply because productivity rises.

Nor does money itself become obsolete. The claim that AI could usher in a “post-money” world confuses the symptoms of scarcity with its underlying causes. Money exists because resources are limited, preferences differ, and tradeoffs must be made. Even in a world of extraordinary abundance, something remains scarce — time, attention, energy, compute capacity, location, or access to networks. Money is the most effective tool yet developed for comparing alternatives and coordinating decisions under such conditions. Remove it, and the allocation problem does not vanish; it reappears in less transparent, often more authoritative forms: queues, quotas, administrative decisions, and political bargaining.

The suggestion that work will become entirely voluntary similarly underestimates the persistence of incentives and institutional constraints. History offers a blunt corrective. Periods of rising material abundance — from the Industrial Revolution onward — have not eliminated work or dissolved economic discipline. Instead, they have shifted the composition of labor, the structure of incentives, and the domains in which effort is applied. At the same time, societies have repeatedly adopted policies and institutions that are inefficient, extractive, or self-defeating, even in the presence of expanding productive capacity. Power, ideology, and incentives do not yield automatically to technology.

This is not to dismiss the transformative potential of AI. On the contrary, the likely expansion of productive capacity is profound. Costs may fall across wide swaths of the economy. New forms of output will emerge. Certain constraints will loosen. But the central economic problem, how to coordinate scarce resources across competing uses over time, does not disappear. It evolves. And with it, the importance of reliable price signals, disciplined capital allocation, and institutional frameworks that align incentives with underlying realities.

Musk’s own business strategy, in fact, points to this conclusion. The same figure who envisions a world of fading scarcity is building and financing some of the most capital-intensive enterprises ever created. These firms depend on scarce inputs: energy, specialized labor, advanced semiconductors, and massive infrastructure. Capital markets exist precisely to allocate those scarce resources based on expectations of future returns.

If abundance were truly imminent and complete, the logic of valuation itself would begin to erode. Investors purchase claims on future profits. A world in which goods and services are universally abundant and prices broadly collapse is one in which sustaining those profits becomes far more difficult. The persistence of large-scale investment, valuation, and competition suggests that scarcity, while shifting, is not disappearing.

The more coherent interpretation of the AI future is therefore not one of total abundance, but of uneven and evolving constraints. Some goods will become dramatically cheaper. Other goods, and many services, are likely to become more valuable precisely because they cannot be replicated at scale. Bottlenecks will emerge in new places; indeed, they are already starting to: in energy grids, data infrastructure, regulatory systems, human expertise. In that environment, money does not lose relevance. It follows scarcity wherever it appears.

The appeal of the post-scarcity narrative is understandable. It offers a vision in which economic tensions dissolve, and policy becomes a matter of distribution rather than discipline. But that vision rests on assumptions that collapse under closer inspection. Technology can expand the frontier of what is possible and available. It does not eliminate the need to choose among possibilities and evaluate trade-offs. It does not erase the importance of how resources are allocated, how incentives are structured, or how information is conveyed through prices. If anything, the arrival of powerful new technologies makes sound economics more important — not less.

READ MORE from Peter C. Earle:

Rare Earths, Real Bottlenecks, and Misguided Policy

Cuba on the Cusp: Getting the Story Right This Time

Stock Market Wisdom, and Its Limits

Image licensed under Attribution-ShareAlike 4.0 International.