What Jobs Can AI Learn? Measuring Exposure by Reinforcement Learning

TL;DR

This paper introduces an RL Feasibility Index to measure which US occupations AI can learn, revealing significant differences from existing AI exposure metrics and informing policy decisions.

Contribution

It develops a novel RL-based approach to assess AI learnability across occupations, providing a more accurate measure of AI's potential impact on jobs.

Findings

RL feasibility scores differ from existing AI exposure measures for certain occupations.

Power plant operators and similar roles score high on RL feasibility but low on traditional AI exposure.

Creative and interpersonal roles show high AI exposure but low RL feasibility.

Abstract

Which jobs can AI learn to do? We examine this for every occupation in the US economy. Existing indices measure the overlap between AI capabilities and occupational tasks rather than which tasks AI systems can learn to perform, and as a result misclassify occupations where the gap between present capability and learnability is large. Reinforcement learning in post-training, now the dominant paradigm at the frontier, is structured around task completion and maps more directly onto the task-based architecture of occupational classifications than prior approaches. Using LLM annotators guided by a rubric developed with RL experts and validated against confirmed deployment cases, we score all 17,951 ONET tasks for training feasibility and aggregate to the occupation level, producing an RL Feasibility Index. The index diverges sharply from existing AI exposure measures for specific occupation groups: power plant operators, railroad conductors, and aircraft cargo handling supervisors score high on RL feasibility but low on general AI exposure, while creative and interpersonal roles (musicians, physicians, natural sciences managers) show the reverse. These divergences carry direct implications for policy interventions.