More pessimistic than Citrini: the more advanced AI efficiency becomes in the US, the more the allocation system breaks down.

On February 23, Citrini Research published “The Global Intelligent Crisis of 2028,” a thought experiment projecting an AI-driven surplus of intelligence between 2026 and 2028. The report does not provide direct predictions but instead uses structured scenario analysis to depict a possible negative feedback loop: continuous improvement of AI capabilities → replacement of white-collar jobs → decline in labor income → contraction of consumption → pressure on corporate profits → further AI investments by companies → larger-scale replacements and layoffs, ultimately affecting private credit, insurance funds, and the high-quality mortgage system.

After the report was released, it triggered a new wave of AI panic trading in the market. U.S. stocks continued to decline, with sectors highlighted in the Citrini report—including food delivery, software, payments, and private credit—experiencing significant drops. Coupled with Nassim Taleb’s market warnings and statements from AI startup Anthropic, IBM’s stock price plummeted 13% in a single day; DoorDash, American Express, KKR, and Blackstone all fell more than 6%; related software ETFs declined 4.8%, widening the total decline since September last year’s peak to about 35%.

It is worth emphasizing that the report does not offer new insights but systematically consolidates several core issues discussed in the market over the past year: AI replacing white-collar workers, worsening marginal returns of SaaS business models, exposure of private credit leverage, and loosening income assumptions for high-quality mortgages. What truly triggered market panic was the framing of these issues as a closed-loop system, where each hypothetical scenario and transmission link can be identified with early signals in reality. The market continues the previous AI panic trading sentiment and falls into a “shoot first, ask questions later” selling pattern.

This round of trading logic has shifted from the previous focus on “how AI will disrupt corporate profit models” to a more macro question: when labor income continues to decline and consumption shrinks accordingly, who will absorb the ever-expanding output? This essentially questions the fundamental distribution mechanisms and the capacity of institutional frameworks.

Increased production efficiency no longer benefits ordinary workers, and asset pricing is being reconstructed