By RICARDO HAUSMANN & ANDRES VELASCO
Is the artificial intelligence boom a bubble? No one can be sure. But one way to answer that question is to ask a more manageable (and more interesting) one: What kind of world economy would have to emerge for today’s market valuations to make sense?
Consider the core group of firms at the center of the AI story: Nvidia, Alphabet, Apple, Microsoft, Meta, Broadcom, Tesla, OpenAI, Anthropic, SpaceX-xAI, and Amazon Web Services. Taken together, they embody a remarkable market wager. Under a conservative benchmark — that by 2036 these firms trade at price-earnings ratios of 20, earn net profit margins of 20 percent and obtain 65 percent of their incremental revenue from abroad — in a decade they would generate roughly $2.4 trillion in additional annual foreign revenue. Such revenue is roughly equal to all US goods exports today and over twice the US current account deficit.
Granted, the US will have to import some inputs (like semiconductors) to provide those AI services. And not all rents will accrue to Americans because foreigners also hold shares in US tech companies. But the offset is likely to be small: foreigners reportedly own just 15 percent to 20 percent of the S&P 500 stock index.
These striking figures should prompt us to rethink much of today’s global macroeconomic debate. For years, discussion of global imbalances has revolved around a familiar concern: how long can the US continue to run large external deficits?
But if markets are even approximately right about AI, the more urgent question is how the rest of the world will pay for the growing claims of US-owned AI capital on global income.
This is a striking reversal. The world will not merely be asked to recognize America’s technological lead. It will be asked to pay for it — year after year and on a vast scale.
Forget the broad-based industrial export surge President Donald Trump keeps promising; the world’s payments will be to a relatively small group of firms that control the large language models, chips, cloud infrastructure, software ecosystems and complementary platforms on which the AI age depends.
How exactly is the rest of the world supposed to pay?
Countries pay for imports with exports of goods and services, investment income, asset sales or borrowing. But the US is moving toward greater protectionism, making it harder for foreign economies to earn the dollars needed to buy American goods and services.
That creates an obvious contradiction. The US cannot expect the world to transfer vast sums to its AI champions while also restricting the channels through which the rest acquire purchasing power.
This tension has far-reaching political implications. Silicon Valley and Trump’s “Make America Great Again” movement may be part of the same domestic political coalition, but their interests are fundamentally misaligned. If the valuation of America’s leading AI firms depends heavily on future foreign revenue growth, then, for them, access to foreign markets is not incidental — it is central. And foreigners who are expected to buy American AI products and services at unprecedented scale must also be allowed to sell something in return.
That is not the end of the story: a massive expansion in American export supply can be accommodated only if the relative price of those exports drops — that is, if the terms of trade of the rest of the world improve vis-a-vis the US. That makes the political trade-off faced by the MAGA coalition even less appetizing.
Seen from this perspective, much of today’s trade debate already looks outdated. The obsession with bilateral deficits and tariff schedules seems oddly parochial — and increasingly irrelevant — when set against the prospect of trillions of dollars in recurring payments to a handful of American technology firms. If current valuations are broadly right, the next phase of globalization will center not on the US trade deficit but on the world’s need to pay for access to US-owned AI infrastructure.
Then there is the question of taxation.
The rents implied by these valuations are enormous. And they accrue to firms that,
taken together, today employ fewer than 1 million people. By this arithmetic, they hold $23 million in market value per worker. This is not a story of broad-based job creation. It is a story of claims by a small group of people on the future income of the rest of humanity.
That makes taxation not just possible but politically indispensable. The US government will want a share. State and local governments hosting data centers will want a share.
And foreign governments will want a share. The current disputes over digital services taxes may come to look like a mild prelude to a much larger struggle over who may tax AI rents.
And taxation matters directly for valuation. If these rents remain lightly taxed, then current prices may be vindicated. But if governments do what governments usually do when confronted with large, visible and politically vulnerable rents, then one of two things must occur. Either these firms will have to generate even more pre-tax revenue and profit than we have assumed or shareholders will ultimately receive less than current valuations imply. In the second case, the bubble interpretation becomes harder to dismiss.
There are also broader geopolitical implications. The rest of the world is unlikely to passively accept an arrangement implying very large recurring payments to a small number of US firms. Countries will try to reduce their dependence on America. They will subsidize domestic alternatives, impose local hosting requirements, favor national champions in procurement, tighten competition policy, and devise new taxes and regulations aimed at clawing back some of the rents. If AI valuations are correct, we can expect not just an era of US corporate dominance but also of mounting resistance to it.
Far from signaling the end of American power, this scenario points to a new form of it. In the 20th century, US power rested heavily on manufacturing scale, military reach and dollar strength. In the 21st, it may rest increasingly on ownership of indispensable AI infrastructure. The challenge for the world will be how to pay for access to it. The challenge for the US will be to sell that infrastructure’s services worldwide in the face of nationalist trade policies and the inevitable pressure to tax extraordinary rents.
Perhaps today’s valuations are a bubble. Perhaps markets are broadly right about the technology but wrong about the politics and the taxation. Or perhaps they are correctly anticipating an economic order in which a handful of US firms captures an unprecedented share of global income. But if the latter scenario is what the future holds, then our public debate is focused on the wrong questions. The defining issues of the next decade will not be tariffs, deficits or American decline. They will be how the world rewards US-based AI capital, how long protectionism can survive that reality and who gets to tax the rents.
• Ricardo Hausmann, a former minister of planning of Venezuela and former chief economist at the Inter-American Development Bank, is a professor at Harvard Kennedy School and Director of the Harvard Growth Lab.
• Andres Velasco, a former finance minister of Chile, is Dean of the School of Public Policy at the London School of Economics and Political Science.
