‘An Unstoppable Feedback Cycle’: How an AI Apocalypse Report Disturbed US Markets

1. AI agents remove all ‘friction’ in the economy

The advent of AI agents has led to a significant leap in their capabilities. This is not merely hypothetical; it reflects ongoing trends in AI technology today. Notable examples, such as Anthropic’s Claude Code and OpenAI’s Codex, have demonstrated remarkable performance, impressing users and experts alike.

As these AI agents advance, they threaten the existence of various software-as-a-service companies, including notable platforms like Monday.com, Zapier, and Asana. By providing businesses with more efficient in-house task management—such as the dealings with databases and workflow organization—these AI solutions offer a more cost-effective alternative. The pressure this creates forces established enterprises like Oracle, reliant on long-term contracts, into a downward spiral of competitive pricing.

In a further twist, consumers are anticipated to leverage personal AI agents to carry out transactions, effectively bypassing traditional service providers that thrive on what can be described as ‘friction’ in economic processes. This shift jeopardizes businesses like travel agencies and real estate firms, which have historically acted as intermediaries in areas like holiday booking and property transactions.

For example, instead of using services like DoorDash for food delivery, both developers and everyday consumers are expected to create their own applications, leading to a fragmented market that adversely affects traditional firms. This phenomenon isn’t isolated to food delivery; companies like Uber also lose their foothold as consumers navigate directly with their own applications. Additionally, the shift of preference towards cryptocurrency for transactions, driven by AI agents, destabilizes longstanding financial institutions such as Visa and Mastercard due to reduced transaction costs.

Citrini argues that these developments represent a natural consequence of AI agents’ inexhaustible capacity to optimize tasks. The very notion of habitual app loyalty—the cornerstone of these businesses—simply evaporates when machines operate, unaffected by personal preferences or brand allegiance.

In market movements reflecting this scenario, the shares of Uber, DoorDash, and financial card companies like Mastercard and American Express saw notable declines recently.

2. Mass white-collar unemployment

Typically, technological advancements create new job opportunities, filling roles left vacant by previous innovations. However, this pattern diverges with the rise of AI.

Citrini notes the unique aspect of AI as it evolves into a general intelligence. This new form of intelligence not only replaces jobs traditionally held by humans but also extends beyond mere operational roles—forcing displaced workers into roles in the gig economy, where stability is often lacking.

Illustrating this trend, Citrini mentions a fictional friend who, after being dismissed from Salesforce, winds up driving for Uber. Such transitions serve to lower wage levels within those already precarious job markets. As these layoffs accumulate, consumer spending inevitably falters, prompting companies to seek even more AI solutions rather than invest in their human capital.

This cycle develops into a self-reinforcing feedback loop devoid of any natural mitigating elements, as it severely impacts the financial clout of the wealthiest segment of the population, which contributes significantly to overall consumer spending.

3. Ripples out into the broader economy

The effects of job losses and the decline of software companies extend far beyond their immediate sectors, resulting in significant impact across broader markets through both defaults in private credit and mortgage crises.

Private credit firms, which do not operate as traditional banks, have played a role in restructuring software firms by extending loans based on predicted future revenues. An illustrative case is the acquisition of Zendesk by Hellman & Friedman and Permira in 2022 for $10.2 billion, contingent on stable revenue projections.

With the emergence of AI agents, such forecasts are rendered obsolete, leading to what could become the largest default in private credit history centered on software. Although initially confined to the tech sector, the repercussions extend to asset managers’ balance sheets—comprising both life insurance policies and household savings.

As regulators downgrade this software debt, the implications contribute to a broader economic downturn by 2027.

In tandem, a mortgage crisis emerges as white-collar professionals lose their jobs and fall short on mortgage repayments, having previously borrowed based on an optimistic future that is no longer attainable.

4. Downward spirals

All of these developments exacerbate existing negative feedback loops.

The first-order spiral manifests as companies lay off workers, which diminishes demand and consumer spending, subsequently prompting further investments in AI and continued layoffs.

A second-order spiral emerges from private credit instability and mortgage concerns, leading to tighter market conditions, a breakdown of consumer confidence, increasing layoffs, and rising mortgage defaults. “Each reinforces the other,” Citrini emphasizes.

Regrettably, traditional financial policy tools offer little recourse, as the crisis is fundamentally rooted in real economic factors—job losses and suppressed wages—rather than financial strains that central banks can readily address. Here, the investment in AI, which effectively reduces the rarity and value of human intelligence, adds to the complexity of the situation.

Consequently, by late 2027, mounting pressures culminate in a crash spurred by perilous mortgage markets, obliterating 57% of the S&P index.

5. Occupy Silicon Valley and Ghost GDP

In the wake of the crash, Citrini theorizes that governments will find themselves incapable of managing the ensuing crises.

As Citrini articulates, “The system wasn’t designed for a crisis like this,” pointing out that typical government revenue relies heavily on taxes derived from human labor and corporate payments.

As tax revenues dwindle amid soaring demands for financial aid, a troubling contradiction surfaces. In contrast, AI corporations thrive, with major tech firms raking in profits from AI innovations, painting an increasingly positive economic picture on paper.

Citrini coins the term “ghost GDP” to describe this phenomenon, where apparent economic output does not circulate through real-world economic channels.

As social tensions rise, movements reminiscent of Occupy Wall Street begin to mobilize, applying pressure to AI firms with prolonged protests.

Concluding this scenario, Citrini cautions, “This is the first time in history that the most productive asset in the economy has produced fewer jobs rather than more.” The necessity for new frameworks to sustain wellbeing amidst these shifts becomes paramount, and the pressing question remains whether society can construct these frameworks in time.

Reactions to the Citrini scenario have been varied. Some analysts, including Stephen Innes, managing partner at SPI Asset Management, observe that current AI discourse has taken on an outsized role in market dynamics, noting that while the market has absorbed numerous challenges, a widely circulated commentary can sway market sentiments with alarming ease.