Descriptions:
A new National Bureau of Economic Research working paper introduces a novel “adaptive capacity” index to reframe the AI job displacement debate—arguing that knowing which jobs face the highest AI exposure is only half the picture. The study analyzes roughly 350 occupations representing 96% of US employment and combines exposure data with a four-factor adaptive capacity score built from liquid savings, age, geographic density, and skill transferability.
The findings produce four worker quadrants. The most concerning: 6.1 million workers face both high AI exposure and low adaptive capacity, with 86% of them being women concentrated in administrative and clerical roles. These workers typically have modest savings, limited skill transferability, and narrower reemployment networks. Geographic vulnerability is also spatially clustered—college towns and state capitals including Laramie, Wyoming, Stillwater, Oklahoma, Springfield, Illinois, and Carson City, Nevada have 5–7% of their local workforce in this high-risk category. By contrast, the 26.5 million workers in high-exposure but high-adaptive-capacity roles—software developers, financial managers, and lawyers—are generally well-positioned to manage transitions.
The host praises the framework’s nuance but raises a foundational critique: every component of the adaptive capacity index was calibrated in an economy where comparable alternative jobs exist. If AI drives structural rather than cyclical displacement, the historical reemployment relationships underlying the model may not hold—a limitation the paper acknowledges only briefly in its caveats section.
📺 Source: The AI Daily Brief: Artificial Intelligence News · Published January 26, 2026
🏷️ Format: Deep Dive
