
The theoretical debate over artificial intelligence’s impact on the economy has abruptly shifted from abstract academic discussion to a visceral market reality. In a defining week for the technology sector, a convergence of viral research essays and high-profile corporate restructuring has ignited what Wall Street is calling the "AI scare trade"—a rush to divest from software and service companies perceived as vulnerable to autonomous displacement.
For years, the consensus narrative was that AI would augment human productivity, creating a "copilot" dynamic that boosted output without eliminating the pilot. However, the events of late February 2026 have shattered that optimism. Fueled by a viral macroeconomic thesis from Citrini Research and a stark warning from AI insider Matt Shumer, investors are now pricing in a "white-collar recession" where autonomous agents don't just assist workers but render them—and the software subscriptions they use—obsolete.
The panic began with the release of two distinct but complementary pieces of writing that captured the anxieties of the moment. The first, a visceral post by AI founder Matt Shumer titled Something Big Is Happening, compared the current state of AI advancement to February 2020—the quiet weeks before the COVID-19 pandemic altered global society. Shumer argued that the release of models like GPT-5.3 Codex and Opus 4.6 marked a "discontinuity" in capability, where AI systems began demonstrating genuine autonomy, self-correction, and "taste."
Shumer’s central thesis was that the gap between model capability and public perception had widened dangerously. He warned that while the general public saw chatbots, industry insiders were witnessing agents capable of displacing entire technical workflows. "We are not making predictions," Shumer wrote. "We are telling you what has already occurred in our own jobs."
Simultaneously, a more detailed and financially devastating argument emerged from Citrini Research. In a 7,000-word report titled The 2028 Global Intelligence Crisis, the firm outlined a hypothetical retrospective from the future, describing a "deflationary cascade" triggered by AI.
Key Concepts from the Citrini Report:
| Concept | Definition | Economic Impact |
|---|---|---|
| Ghost GDP | Economic output that rises due to automated productivity but fails to circulate as wages. | Creates a disconnect where stock valuations rise while consumer spending power collapses. |
| The Seat-Based Cliff | The collapse of software revenue models that rely on "per-user" licensing. | As AI agents replace human workers, companies cancel thousands of SaaS subscriptions (e.g., Salesforce, Zoom). |
| Intelligence Displacement Spiral | A feedback loop where companies use AI to cut costs, weakening demand, forcing further AI cuts. | Results in a structural decline in white-collar employment and creditworthiness. |
| The Prime Borrower Crisis | The insolvency of high-earning professionals who hold the bulk of mortgage and credit debt. | Threatens the stability of housing markets and premium credit card issuers like Amex. |
The Citrini report resonated because it provided the macroeconomic framework for Shumer’s technical warnings. It coined the term "Ghost GDP" to describe a future where corporate profits soar due to automated efficiency, but the velocity of money collapses because machines do not buy houses, go to dinner, or pay income tax.
The immediate victim of this narrative shift was the software-as-a-service (SaaS) sector. For a decade, the "per-seat" business model—charging $20 or $50 per employee per month—was the gold standard of tech investing. The logic of the AI scare trade is simple: if AI agents replace human employees, there are fewer "seats" to monetize.
Market reaction was swift and brutal. Stocks of major productivity platforms, including Datadog, Salesforce, and even previous darlings like CrowdStrike, saw double-digit percentage drops. Investors began to view these companies not as beneficiaries of AI, but as victims of a shrinking user base. If a marketing department of 50 people is reduced to three managers overseeing a swarm of autonomous agents, the software revenue from that department collapses by 94%, even if productivity remains constant.
This "deflationary cascade" challenges the foundational valuation metrics of the modern tech economy. It suggests that the efficiency gains from AI will be captured almost entirely by the companies deploying the AI (or the hyperscalers providing the compute), rather than the intermediate software vendors who rely on human headcount.
If the essays provided the theory, Jack Dorsey provided the proof. Amidst the swirling debate, Dorsey’s fintech company, Block (formerly Square), announced a massive 40% reduction in its workforce. Unlike traditional layoffs blamed on "macroeconomic headwinds" or "overhiring," Dorsey explicitly cited the transformative efficiency of internal AI tools as the primary driver.
The announcement served as a grim validation of the Citrini thesis. It demonstrated that the "replacement" phase of AI adoption was no longer hypothetical. Companies are beginning to realize that "agentic workflows"—where AI systems plan and execute complex tasks without human intervention—allow for leaner operations than previously thought possible.
Sectors Most Vulnerable to the "AI Scare" Trade:
| Sector | Vulnerability Factor | Representative Stocks Hit |
|---|---|---|
| Enterprise SaaS | Revenue tied directly to human headcount (per-seat pricing). | Salesforce, Workday, Atlassian |
| Business Process Outsourcing (BPO) | Tasks (call centers, data entry) are easily fully automated. | TaskUs, Teleperformance, Genpact |
| Consulting & Professional Services | Junior analyst work is replaced by LLM reasoning. | Accenture, McKinsey (Private), Legal Services |
| Creative Agencies | Generative media replaces human design and copy labor. | Omnicom, WPP, Adobe (Mixed impact) |
The Block layoffs sent a chill through the white-collar workforce. It signaled that the "jobless growth" predicted by the Ghost GDP theory might be arriving sooner than policymakers expected. The fear is that Block is merely the first domino, with other major tech and finance firms secretly preparing similar restructuring plans to capitalize on the new "unit economics of intelligence."
The term "Global Intelligence Crisis," coined by the Citrini report, frames the abundance of intelligence not as a utopia, but as a monetary shock. In this scenario, intelligence becomes so cheap that it acts as a deflationary force on human labor capital.
The report argues that the U.S. economy is structured around the consumption habits of the white-collar middle class. This demographic holds the mortgages, buys the new cars, and drives discretionary spending. If their labor value drops to zero—or if they are forced into lower-paying "human-in-the-loop" roles—the ripple effects could destabilize the housing market and the credit system.
Critics of the scare trade, including some Federal Reserve governors, have pushed back. Fed Governor Christopher Waller labeled the reaction "overstated," arguing that AI is a tool that will lead to redeployment rather than permanent displacement. They point to historical precedents where technology created more jobs than it destroyed.
However, proponents of the scare trade argue that autonomous agents represent a fundamental break from history. Unlike the steam engine or the spreadsheet, these agents possess "general" capabilities. They don't just speed up a task; they perform the cognitive labor of deciding what task to do next. As Shumer noted in his essay, when the tool becomes capable of "taste" and "judgment," the human role of "manager" is the only one left—and there is only room for so many managers.
As we move further into 2026, the tension between the "AI Boom" (investing in hardware/infrastructure) and the "AI Scare" (selling the software/labor layer) will define the market. The "Ghost GDP" phenomenon poses a unique challenge for the Federal Reserve: how do you manage monetary policy in an economy with skyrocketing productivity but potentially stagnating or falling employment?
For the tech industry, the message is clear: the era of "selling tools to humans" is ending. The new era is about "selling outcomes to enterprises." Companies that charge for work done (e.g., resolving a customer support ticket, closing a sale, writing a codebase) rather than seats filled may survive the transition.
The events of this week have forced a collective reckoning. The doomsday essays are no longer just science fiction; they are being read in boardrooms as potential roadmaps. Whether this leads to a "crisis" or a "renaissance" depends on how quickly society—and the market—can adapt to a world where intelligence is abundant, but human labor is no longer the scarce resource it once was.