Citrini’s Scenario Is A Great But Deeply Flawed Thought Experiment

A speculative economic scenario by 'Citrini' has gone viral, with Bloomberg noting it contributed to a decline in the stock market. The essay, analyzed in depth by Zvi on LessWrong, presents a concrete narrative where rapid AI advancement from 2026-2028 leads to widespread white-collar job displacement without corresponding economic stimulation. The scenario culminates in a 38% drawdown in the S&P 500 from its peak and unemployment hitting 10.2% by June 2028, describing an economy that 'no longer resembles the one any of us grew up in.'

The piece is praised as a valuable 'thought experiment' that introduces underexplored economic mechanisms and concrete numbers, labeled upfront as 'a scenario, not a prediction.' However, Zvi's critique highlights severe flaws: the pace of AI capability diffusion is unrealistically fast ('Can't Happen levels'), it ignores the monumental implications of the superintelligence it presupposes ('we probably all die'), and it assumes a passive government response. The scenario also controversially posits that the NASDAQ would only rise 32% from current levels before the crash, a figure critics find surprisingly low given the premise of a transformative AI boom.

This viral discussion matters because it frames AI's economic 'tail risks' not as job apocalypses or rogue agents, but as a deflationary cascade where AI destroys margins and jobs without creating new demand. It forces economists and tech professionals to model second-order effects, even if the specific narrative is flawed. The debate exemplifies Cunningham's Law in action—posting a provocative 'wrong' answer to generate better models of AI's real-world economic impact.