Labor market impacts of AI: A new measure and early evidence

"We introduce a new measure of AI displacement risk, observed exposure, that combines theoretical LLM capability and real-world usage data, weighting automated (rather than augmentative) and work-related uses more heavily. AI is far from reaching its theoretical capability: actual coverage remains a fraction of what's feasible."

"Occupations with higher observed exposure are projected by the BLS to grow less through 2034. Workers in the most exposed professions are more likely to be older, female, more educated, and higher-paid, suggesting differential vulnerability across demographic groups."

"We find no systematic increase in unemployment for highly exposed workers since late 2022, though we find suggestive evidence that hiring of younger workers has slowed in exposed occupations, indicating potential labor market adjustments."

Researchers introduce 'observed exposure,' a measure combining theoretical LLM capabilities with actual usage patterns to assess AI displacement risk, weighting automated and work-related applications more heavily. Current AI adoption remains far below theoretical potential. Occupations with higher observed exposure show slower projected employment growth through 2034 according to BLS forecasts. Workers in highly exposed professions tend to be older, female, more educated, and higher-paid. Despite these patterns, no systematic increase in unemployment has occurred among highly exposed workers since late 2022, though evidence suggests hiring of younger workers may have slowed in exposed occupations. The research acknowledges historical limitations in predicting technological labor market impacts.

Read at Anthropic