BlockBeats report, March 7: Anthropic’s latest research shows that although AI theoretically covers the majority of tasks in fields such as business, finance, law, and computer science, actual adoption rates remain low—using the Claude model as an example, its theoretical coverage for computer and mathematics roles is 94%, yet actual usage stands at only 33%. The study introduces a metric called “observed exposure” to compare theoretical capabilities with real-world usage data.
The results show that the group most exposed to AI is not blue-collar workers, but highly educated, high-income female white-collar workers: this group has a 16-percentage-point higher proportion of women, 47% higher average income, and nearly four times the proportion of individuals with graduate degrees compared to the low-exposure group. Researchers warn that as AI capabilities improve and adoption deepens, it could trigger a “great recession for white-collar workers”—similar to how unemployment doubled from 5% to 10% during the 2007–2009 financial crisis.
Although not yet realized, the risk is clear. The current impact is more evident in slowed hiring rather than layoffs: since the ChatGPT era, job search rates for vulnerable professions have dropped by 14% compared to 2022, and employment rates among young workers aged 22–25 in related fields have declined by 16%. Some young people are choosing to pursue further education or temporarily exit the labor market.