An AI Doomsday Memo Rattled Markets as Pressure Builds on White-Collar Work
US markets do not normally move because of a speculative essay, yet in late February a long AI scenario published on Substack managed to do exactly that. Shares in several platform and payment companies slid sharply, including Uber, Mastercard, American Express and DoorDash, while the broader S&P 500 dipped as investors digested a dramatic thought experiment about how artificial intelligence might ripple through the American economy.
Work News | New Stardom
Trend Analysis
Photo by Igor Omilaev ‍
The essay came from Citrini Research, a small macro analysis firm that writes about long-term technological and economic shifts. The report, titled The 2028 Global Intelligence Crisis, was explicitly framed as a scenario rather than a prediction. Written from the fictional perspective of June 2028, it imagines a near future in which increasingly capable AI agents automate large parts of the digital economy and trigger a cascade of economic effects.
In the Citrini scenario the disruption begins inside software. AI systems become capable enough to handle many of the tasks currently performed by engineers and analysts, allowing companies to generate tools internally rather than paying for expensive software subscriptions. As that pressure spreads across the software-as-a-service sector, pricing collapses and layoffs follow. The consequences then extend beyond technology companies themselves. White-collar job losses reduce spending by high-earning households, mortgage repayments weaken, and financial stress spreads through private credit markets tied to software firms.
The story ends with a full-scale economic shock. By the end of the fictional timeline unemployment rises above ten percent and the S&P 500 falls dramatically as mortgage defaults spread through the financial system. The most provocative part of the argument is not the market crash itself but the logic behind it: if AI significantly reduces the demand for high-income knowledge work, the spending power that drives large parts of the modern economy weakens with it.
That argument travelled fast because it arrived in a moment when investors were already trying to price the economic impact of AI. Within days a counterargument appeared from Frank Flight, a macro strategist at Citadel Securities. His response, titled The 2026 Global Intelligence Crisis, took a very different view of the same question. Rather than extrapolating from AI capability trends, the Citadel analysis focused on current economic data. Software hiring remains strong, unemployment is still low and new business formation in the United States continues to rise.
In this view AI is best understood as a productivity shock that will spread gradually through the economy, following the pattern seen with previous technologies such as personal computers and the internet. Companies do not reorganise overnight. Infrastructure takes time to build, enterprise procurement cycles move slowly and organisational change tends to lag far behind technological possibility. Even if AI eventually transforms many jobs, Citadel argues that the transition is unlikely to unfold at the speed imagined in the doomsday scenario.
Both perspectives agree on one point: artificial intelligence will reshape the composition of work. The disagreement lies in how quickly that process might unfold and how disruptive it will be along the way. The reason the Citrini memo unsettled markets is that it pushes a question that has become harder to ignore. For decades automation primarily threatened routine or manual labour. AI systems are now beginning to perform tasks that sit at the core of white-collar work.
That shift explains why the debate resonates far beyond trading desks. A growing share of professional work consists of searching, summarising, drafting, analysing information and coordinating digital processes. These tasks once required large numbers of well-paid employees. Increasingly they can be performed by software that operates continuously and at very low marginal cost. The result is not necessarily immediate job elimination but a steady erosion of the economic scarcity that previously protected many office roles.
Even if the most extreme forecasts prove exaggerated, the direction of travel is difficult to miss. Teams are expected to produce more output with fewer people. Junior roles that once served as training grounds for professional careers are becoming harder to justify when software can perform entry-level research, coding or documentation work in seconds. At the same time the gap is widening between workers who know how to integrate AI into their workflows and those who still treat the technology as an occasional convenience.
For white-collar professionals the lesson is not that an economic collapse is inevitable, but that waiting for clarity is a poor strategy. The safest assumption is that generic knowledge work will become less valuable unless it is combined with capabilities that machines struggle to replicate or unless the worker themselves becomes highly effective at using those machines.
That means treating AI tools less like curiosities and more like infrastructure. Workers who understand how to build real workflows with AI systems, automate parts of their own tasks, evaluate outputs critically and combine these tools with existing software are already becoming more productive inside the same roles. The difference between experimenting with AI and relying on it daily is quickly becoming visible inside organisations.
At the same time it is increasingly risky to rely on a single narrow form of knowledge work as the sole source of income. The modern services economy rewards people who combine technical literacy with domain expertise, industry knowledge or strong human networks. Judgement, negotiation, leadership, regulatory understanding and the ability to operate in messy real-world environments remain harder to automate than routine digital analysis.
Finally, professionals who want to understand where the economy is heading would do well to watch the infrastructure behind the AI boom itself. Data centres, energy projects and cloud infrastructure are absorbing hundreds of billions of dollars in investment. Those projects move slowly and require years of planning, but they also reveal where companies believe future demand will come from.
The Substack essay that rattled markets may turn out to be an exaggerated thought experiment. Yet the speed with which it travelled says something important about the moment. Investors, executives and workers are all trying to understand what happens when machines begin competing directly with the most expensive forms of human labour. Whether the transition unfolds over two years or twenty, the pressure building around white-collar work is no longer hypothetical.
Follow global work and job trends. Subscribe to New Stardom’s newsletter.
This work is funded directly by readers who value independent reporting.
If you find this useful, you can support New Stardom’s independent journalism.
Read Next
Have insights on work and the future of work? Submit an opinion piece to New Stardom. Love work and career books? Explore our fun workplace book collection.
New Stardom is an independent magazine covering the Future of Work, AI, and emerging job trends.
