Author: Chasing Wind Trading Platform
When discussing AI, most people are still worried about “jobs being taken away.” But Deutsche Bank believes this perspective might be a bit narrow.
According to Chasing Wind Trading, in a recent report by George Saravelos, Head of Global Foreign Exchange Research at Deutsche Bank, two extreme scenarios of AI development are explored:
First scenario: “Complete Substitution.” Like Marx’s prophecy over 180 years ago and Elon Musk’s current vision: in the economic factors of production, “capital” itself becomes “labor,” rendering labor value zero, and making capitalism obsolete. Large-scale AI replacing human jobs leads to wealth and income being highly concentrated among a few capital owners, while ordinary people’s income and demand weaken, causing an economy to face a paradox of “abundant goods, but no one can afford them.”
Did Marx predict artificial intelligence? About 200 years ago, he wrote a book about “machines,” imagining a fully automated world. In this scenario, scarcity issues are resolved. But as labor value drops to zero, capitalism becomes outdated, and we transition into a new world of material abundance. Marx’s envisioned endpoint is eerily similar to Elon Musk’s current vision.
Second scenario: “History Repeats.” AI, like previous technological revolutions, improves efficiency but does not fully replace human labor; instead, it “empowers” humans, with new jobs constantly emerging, and policy systems able to mitigate shocks. In this case, the economic logic remains similar to the past decades, with inflation, interest rates, and stock markets likely experiencing moderate growth.
Are we heading into an abyss, a utopia, or just experiencing a normal industry upgrade? This report from Deutsche Bank offers us a new perspective.
To understand AI’s ultimate disruptive power on the economy, we must revisit the origins of modern economics.
Starting with Adam Smith, all classical economists based their theories on a fundamental assumption: capital and labor are two completely independent factors of production. Whether it’s capital or labor, their prices (interest rates and wages) are determined by their relative scarcity in the market.
Looking back over the past two centuries, all waves of technological innovation have largely conformed to this model.
By analogy, the invention of the steam engine displaced coachmen but created train drivers; the internet destroyed traditional print media but generated countless programmers and delivery workers. In these historical cycles, labor always has something to do. Machines are capital, but operating, maintaining, and designing machines are still labor. Capital is merely a “complement” to labor.
However, fully autonomous robots with Artificial General Intelligence (AGI) completely break this classification.
“In this scenario, capital becomes labor. It is no longer a complement but a substitute.” Saravelos sharply points out in the report.
When an AI machine can think independently, produce autonomously, and iterate on its own, it is both capital and labor. The foundational structure of modern economics fractures at this moment.
The report bluntly states: “When capital equals labor, the value of work drops to zero, and wages also fall to zero. Economists call this an unacceptable equilibrium. Scientists call it a singularity. Classical economic theory collapses. Consequently, capitalism as a system will also become obsolete.”
What happens to macroeconomic operations when labor is massively replaced? Deutsche Bank delves into deeper theoretical implications.
In a purely “AI replacing workers” world, wages decline, but material abundance skyrockets. Machines tirelessly produce vast quantities of goods and services.
According to classical economists like Say, Walras, and Wicksell, “supply will automatically create its own demand.” Their models assume markets have self-correcting abilities. Prices of goods fall as production costs decrease, workers can buy more with less money, or find new jobs in emerging sectors.
However, Deutsche Bank warns that in a fully automated AI world, this self-correction mechanism will break down entirely.
The logic is straightforward: automation will concentrate wealth and income into a small “capital owner” class. In economic terms, the marginal propensity to consume of the wealthy (capital owners) is much lower than that of ordinary workers.
For example, an AI factory can produce 10,000 cars a day at very low cost. But all these profits go to the AI owner. This owner cannot buy 10,000 cars alone; meanwhile, many ordinary people who lose their jobs and see their income drop to zero cannot afford to buy even cheap cars.
“The transmission chain from supply to demand is broken,” Saravelos writes.
This market equilibrium, characterized by structural low labor income, deflationary prices, and massive “excess savings” replacing strong demand, aligns with the “secular stagnation” scenario proposed by economists Eggertsson and Mehrotra. In extreme cases, it could even trigger a Marxist-style revolution.
Can Keynesian economics, the other pillar of modern economics, rescue the situation?
Keynes’s revolution was recognizing the failure of classical theory. Under Keynesian frameworks, economic downturns are not permanent but cyclical. When prices adjust slowly and re-skilling labor lags, government intervention is necessary.
In the AI era, such intervention might involve: imposing high “AI taxes” on AI companies, creating a fund to distribute “stimulus checks” or universal basic income (UBI). Through strong fiscal transfers, the economy can reach a new equilibrium.
But this logic faces significant practical constraints.
The report cites extensive research by economists like Acemoglu and Johnson on the history of technological deployment. History shows policy and institutional adjustments are often very slow.
For example, during the early British Industrial Revolution, due to lack of institutional protections, workers’ real wages were suppressed for decades.
To prevent a decline in living standards, Deutsche Bank lists necessary reforms: “Stronger labor bargaining institutions, competition policies to curb monopolies, tax and subsidy structures that do not favor capital over labor, public investments in skills and creative tasks, and expanded or reformed corporate governance.”
If technological change outpaces government and institutional adaptation, Keynesian remedies may be ineffective or delayed.
Even with a highly responsive, proactive government, deeper political-economic challenges remain.
The report presents a philosophically profound phenomenon: nearly 200 years ago, Marx’s vision of “machines” and full automation eerily resembles Elon Musk’s ultimate AI vision today.
In this fully automated scenario, humanity solves the ancient problem of scarcity.
But this leads to the disintegration of social consensus: “In this fully automated world, the essence of capitalism collapses. Political debates no longer revolve around how to subsidize wages. Instead, they become more fundamental: if scarcity is eliminated, what is the meaning of property rights?”
As Keynes asked in his 1930 essay “Economic Possibilities for Our Grandchildren”: when humans no longer need to work for survival, what is the purpose of human existence?
Though these topics seem grand, Deutsche Bank emphasizes that, given their existential nature, they are deeply connected to current financial market pricing.
For markets, it’s essential to consider both the “transition period” to the endpoint and the endpoint itself. Deutsche Bank divides the future into two parallel universes and provides clear asset pricing logic.
This is a world where AI can rapidly and almost completely replace human labor. From a living standards perspective, it’s a utopia where scarcity is permanently resolved. But Deutsche Bank warns that reaching this state involves “the most destructive and uncertain path.”
In this scenario, AI does not trigger a singularity but functions as an enhancement technology, similar to past innovations in the 20th century.
Deutsche Bank emphasizes that the report is not about making absolute predictions but about establishing an analytical framework. Given the broad range of possible outcomes, debates on AI’s macroeconomic impact will continue in the short term.
From an investor’s perspective, how should we observe the evolution of the AI economy? Deutsche Bank distills clear “observation markers”:
Related reading: Trading Moment — Tech stocks retreat drags down the market, BTC attempts to recover 70,000 USD but fails