Citrini Research 在二月底引发市场波动的 AI 报告，从第一句就强调那是「情境」，不是预测。其核心链条是：AI 投资推升 AI 能力，压低劳动需求，进而带来更低工资与更高失业；之后消费支出下滑，房贷违约增加，并可能引发股市重挫等连锁反应。

为了把这种广泛叙事转成可判定问题，Kalshi 在二月底推出一个预测市场；若到 2028 年 7 月前，五项条件中至少三项发生，市场即判定为 Yes。五项门槛分别是：美国失业率超过 10%；S&P 500 较推出时下跌超过 30%；Zillow Home Value Index 在 NYC、LA、San Francisco、Chicago、Houston、Phoenix 任一城市年减超过 10%；任一季度劳动所得占国内总所得首发值低于 50%；或任一月份 CPI 低于 0%。

截至 3 月 6 日星期五美东时间 16:00，市场给出的机率是 13%，表示在 2028 年 7 月前至少三项结果同时出现的可能性相对不高。这提供了对 Citrini 情境的一次量化现实检验，但也暴露预测市场的限制：规则只要求结果发生，并不要求 AI 是原因，因此在「可判定性」与决策者关心的因果与质性问题之间，仍存在持续的取舍。

The AI report from Citrini Research that moved markets in late February stated from its first sentence that it was a scenario, not a prediction. Its core chain was: AI investment raises AI capability, reduces labor demand, and then leads to lower wages and higher unemployment; consumer spending then falls, mortgage defaults rise, and a stock-market crash and other spillovers may follow.

To convert that broad narrative into resolvable questions, Kalshi launched a prediction market in late February; it resolves Yes if at least three of five conditions occur before July 2028. The thresholds are: the US unemployment rate exceeds 10%; the S&P 500 falls more than 30% from launch; the Zillow Home Value Index drops more than 10% year over year in NYC, LA, San Francisco, Chicago, Houston, or Phoenix; labor share of gross domestic income in any first release falls below 50%; or monthly CPI falls below 0%.

As of 4 p.m. ET on Friday, March 6, the market implied a 13% chance that at least three outcomes occur by July 2028, suggesting a relatively low probability for the combined scenario. This offers a quantitative reality check on the Citrini case, but it also shows a limit of prediction markets: the rules require only that the outcomes happen, not that AI causes them, leaving a persistent tradeoff between resolvability and the causal, qualitative issues decision makers actually care about.