The Viral Essay That Predicted an AI Apocalypse by 2028

In late February, a Substack essay called "The 2028 Global Intelligence Crisis," written by Citrini Research analyst Alap Shah, went viral in financial and tech circles. The piece described a scenario where AI-driven job displacement hits so fast and so hard that the economy can't adjust: the S&P 500 drops 38%, unemployment spikes to 10.2%, and a deflationary spiral sets in as white-collar workers lose purchasing power en masse. The essay reportedly contributed to a $600 billion single-day drop in equity markets.

It is the kind of scenario that sounds plausible if you don't look too hard at the data. Citadel Securities looked hard at the data.

What Citadel found

In a macro strategy report authored by analyst Frank Flight, Citadel systematically worked through Citrini's claims using real-time labor market indicators. The results were not kind to the doomsday narrative.

On the labor market: Indeed job posting data shows demand for software engineers rising 11% year over year in early 2026, not collapsing. The St. Louis Fed's analysis of the Real-Time Population Survey found that daily use of generative AI for work has remained stable and shows "little evidence of any imminent displacement risk."

On historical precedent: Citadel's report argues that rising productivity has consistently lowered costs and expanded what people can buy, prompting demand to shift into new goods and services rather than simply disappearing. The adjustment isn't always painless. Citadel doesn't claim it is. But the 2028 timeline in Citrini's piece assumes displacement will be faster than any prior technology transition and that the economy will have no adaptive response. Neither assumption holds up well against the historical record.

What this pattern looks like

The "2028 Global Intelligence Crisis" essay belongs to a recurring genre: a vivid, internally consistent scenario about technological catastrophe that feels more rigorous than it is because it includes specific dates and numbers. The S&P dropping exactly 38%. Unemployment at exactly 10.2%. The precision is atmospheric. It signals analysis while substituting narrative for evidence.

That said, Citadel is not a neutral party. Their interest in debunking doomsday scenarios is not purely epistemic. It's worth holding both things at once: the specific predictions in Citrini's essay don't hold up to scrutiny, and the broader concern about AI's pace of disruption is not invented.

Whether AI displaces workers faster than new jobs appear is a genuine empirical question without a final answer yet. An April Fortune piece followed an economist who had been among the most alarmed by AI job displacement and found him cautiously reconsidering. The evidence, he said, isn't pointing where he expected. That's the more honest place to sit than either the viral doomsday scenario or its confident debunking.