Real-World AI Task Usage Across Occupations

We detail a novel empirical study analyzing AI usage patterns by examining millions of Claude.ai conversations through the lens of U.S. Department of Labor's O*NET database of tasks and occupations. The research found that AI use is concentrated in software development and writing tasks, accounting for nearly half of total usage, but also extends to about 36% of occupations for at least a quarter of their tasks. The study differentiates between augmentative AI use (57%), which enhances human capabilities, and automative use (43%), which substitutes for human labor. Notably, AI usage is highest in mid-to-high wage occupations requiring considerable preparation (Job Zone 4) and cognitive skills, showing lower use in roles requiring physical labor or extensive training. While acknowledging limitations, the framework offers an automated, granular approach for tracking AI's evolving role in the economy and identifying potential future impacts, providing empirical data that both aligns with and challenges previous predictions.