1 source filed · EN 1
ONET-Bench is built from 45,796 worker ratings of AI responses to ONET-derived tasks; the test split has 4,501 ratings.
3.0%-97.9%share of responses LLM judges rated acceptable, versus 61.1% for occupation-matched workers
1 source filed · EN 1
Robots can perform 3/4 of physical tasks (34% of US working hours) in limited settings; about 80% of task time exposed to robots or LLMs combined; exposed workers more male, less educated, lower paid.
1 source filed · EN 1
Shares of jobs: 18% higher automation risk, 24% reorganize, 12% grow with AI, 46% less immediate change.
18%share of jobs at relatively higher short-term automation risk