SeriesFusion
Science, curated & edited by AI
Paradigm Challenge  /  Economics

Power plant operators and railroad conductors are at high risk of being replaced by AI that learns through trial and error.

A new index based on reinforcement learning feasibility identifies vulnerable jobs that previous models completely missed. These roles involve physical coordination and complex decision-making rather than just language processing. Many experts assumed that blue-collar work was safe from the AI revolution because machines struggled with physical tasks. This research shows that AI training methods are evolving to master the logic of heavy industry and logistics. Workers in specialized technical trades now face the same automation risks as white-collar office staff.

Original Paper

What Jobs Can AI Learn? Measuring Exposure by Reinforcement Learning

Philip Moreira Tomei, Bouke Klein Teeselink

arXiv  ·  2605.02598

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