The Employment Winter under the AI Wave: A Potential Fuse for the Next Great Depression?

The impact of artificial intelligence on the labor market is triggering a deep debate on Wall Street. Pessimists paint a doomsday scenario of economic collapse triggered by technological unemployment, believing that an employment winter could evolve into the next Great Depression; optimists cite historical patterns to argue that technological revolutions will ultimately create more wealth. The core of the disagreement lies in how quickly AI will penetrate the economy.

A recent report by Citrini Research, "Global Intelligence Crisis 2028," presented as a "memo from the future," outlines an AI-driven economic disaster scenario that has garnered widespread attention in the market. The report posits that AI is compressing software development costs and spreading through the entire economy via a "negative feedback loop"—corporate layoffs, shrinking consumption, and narrowing profits, which leads to the acquisition of more AI capabilities in a cycle with "no natural braking mechanism." Under this narrative, the employment winter is not alarmist talk but an approaching reality.

In response, Citadel Securities and Bianco Research have published articles to refute this, asserting that technological diffusion historically follows an "S-curve" pattern. They contend that the speed and breadth of AI's impact are far less than Citrini assumes, and the labor market, corporations, and governments have sufficient time to adapt, making a Great Depression-style scenario lack historical and empirical basis. The debate among these three parties provides an important reference for investors to understand this potential structural risk.

Citrini's Warning: From SaaS Collapse to Systemic Risk

Citrini's pessimistic narrative begins with the SaaS industry. The report points out that with the help of AI tools like Claude Code or Codex, a capable developer can now replicate the core functions of a mid-tier SaaS product in a matter of weeks. This directly shakes the subscription revenue model that the SaaS industry relies on for survival.

However, Citrini's truly alarming argument is not the decline of software companies themselves, but the resulting vicious economic cycle: Improved AI capabilities → Corporations reduce salary expenses → Weakened consumption → Narrowing profits → Corporations acquire more AI capabilities. The report calls this process the "intelligent substitution spiral"— a large number of white-collar workers are pushed into the gig economy, depressing wage levels and dragging down overall economic activity.

Citrini also points out that consumer spending accounts for 70% of U.S. GDP, while machines, acting as "new employees," have zero consumption of disposable goods. Meanwhile, government finances will face dual pressures: declining tax revenues amid existing large fiscal deficits and debt burdens, coupled with the need to transfer more funds to households.

The report also warns of chain risks in financial markets, including defaults in software-related private credit impacting insurance companies, and the repayment risk facing the approximately $13 trillion U.S. mortgage market—"In 2008, loans were bad from day one; in this scenario, loans are initially of good quality, but the world changes after they are issued."

Citadel's Rebuttal: Three Reasons Why the "Intelligent Substitution Spiral" Won't Happen

Citadel Securities presents a systematic refutation of Citrini's core assumptions.

On the data level, Citadel points out that job postings for software engineers have increased by 11% year-over-year, and tracking data on AI workplace applications from the St. Louis Fed "shows virtually no imminent substitution risk."

On the macro-logic level, Citadel uses the national income accounting identity Y=C+I+G+(X−M) Y=C+I+G+(X−M) to argue: If AI drives productivity growth and promotes real GDP growth, then there must be a corresponding expansion in consumption, investment, government spending, or net exports on the demand side. "An economy cannot achieve increased output simultaneously with decreased sales; basic mathematical logic and capitalist motives do not allow for such a scenario."

On historical precedent, Citadel argues that the adoption of steam power, electrification, the internal combustion engine, and even computers has followed an S-curve of technological diffusion—after an acceleration phase, adoption rates tend to plateau as organizational integration costs rise, regulation intervenes, and marginal returns diminish. Even if Citrini's assessment of AI's long-term impact is accurate, its realization speed will be significantly slower than hypothesized, and this buffer period is sufficient for the labor market, businesses, and governments to complete adjustments.

The report cites Keynes' famous 1930 prophecy as evidence: Keynes once expected that productivity gains would shorten the workweek to 15 hours by the early 21st century. While his judgment on productivity trends was proven correct, his projection for the labor market fell short—"Society did not significantly reduce work; instead, it greatly increased consumption."

Bianco: How Jevons' Paradox Negates the AI Substitution Argument

Bianco approaches the issue from another perspective, with the core argument being that Citrini's "fatal flaw" lies in assuming a finite number of problems for humans to solve.

Bianco invokes Jevons' Paradox, pointing out that when technology makes something more efficient, the demand for that thing often explodes rather than contracts. If AI reduces the cost of drafting legal complaints to near zero, lawyers will not sit idle at home; instead, they will file more lawsuits, which in turn generates new demand for legal defense and judges.

Bianco further distinguishes the key variable of "which part of the work AI is automating." He uses London taxi drivers as an example: GPS automated the scarce skill of "memorizing 25,000 streets," leading to a flood of competitors and flattened wages; whereas computers eliminated repetitive, simple parts of accounting, allowing accountants to focus on higher-value financial consulting, thus increasing their wages. "The outcome in the labor market depends entirely on whether AI automates the scarce, high-judgment parts of work or the repetitive, auxiliary parts."

Bianco believes that AI eliminates simple, repetitive links in knowledge work, thereby making practitioners more valuable. This directly contradicts Citrini's judgment—the latter believes that as long as automation is fast enough and large enough in scale, the labor market cannot absorb the shock, regardless of whether the remaining work is more valuable.

Three Parties' Consensus: Speed is the Key Variable

Despite clear disagreements in their conclusions, the three parties share an implicit consensus on one core issue: the speed of the transition period is everything.

Bianco introduces the historical concept of the "Engels Pause"—during the Industrial Revolution, between 1790 and 1840, large-scale unemployment was not promptly offset by new jobs, leading to the spread of communism globally. Citrini's scenario is essentially a modern version of the "Engels Pause," but faster, and thus posing a greater potential shock.

Bianco and Citadel do not deny transition period risks but believe that as long as the speed of AI adoption is controllable and institutional responses are timely, this gap is manageable. All three implicitly agree: If the rate at which jobs disappear consistently outpaces the rate at which new jobs are created over the long term, even if productivity and corporate profits continue to improve, the political and social consequences will be impossible to ignore.

For investors, the 2028 scenario described by Citrini currently remains a tail risk rather than a baseline scenario. Key indicators to track include the number of white-collar job openings, real wage growth in knowledge-intensive industries, and consumer spending trends among high-income households. If these indicators deteriorate simultaneously, the negative feedback loop depicted by Citrini may transition from a theoretical exercise to a real threat.

Risk Disclosure and Disclaimer

Markets involve risks, and investment requires caution. This article does not constitute personal investment advice, nor has it considered the specific investment objectives, financial situation, or needs of individual users. Users should consider whether any opinion, view, or conclusion in this article is consistent with their specific circumstances. Investment based on this article is at the user's own risk.