TaskNet: AI-Led Worker Interviews Add 79 Tasks Missing from O*NET, Raise AI Exposure for Lawyers and Devs
stanfordnlp · x · 2026-11-11
A Stanford team including Erik Brynjolfsson and Diyi Yang published TaskNet as an NBER working paper: an AI-led interview system that builds occupational task inventories directly from workers' accounts.
- Interviewed 132 workers across six occupations, producing 202 task statements from 2,726 reports — all accepted by ONET Center analysts, vs 65% acceptance for write-in suggestions under the current ONET process
- Covers 75% of the 118 published ONET tasks for sampled occupations, adds 79 tasks with no ONET counterpart, and refines 58; covers 89% of held-out workers' reported tasks (vs 79% for ONET)
- Better alignment with employee job postings (72.5% vs 70.2%)
- Finds higher AI exposure for lawyers and software developers, with implications for workforce policy and training
More from AGI Musings
- AI isn't a tool for the rich — it's the first intelligence lever for everyone else — LucaAmb · 2026-11-11
- Anthropic CEO predicts cancer and Alzheimer's cured in under 10 years — ns123abc · 2026-11-11
- AI agents cut resources to break ECDSA signatures by 90%, beating Google's estimate — IgorCarron · 2026-11-11
- Stanford scholars launch nonprofit STABL to build societal resilience against advanced AI — RishiBommasani · 2026-11-11
- METR report: AI-driven R&D acceleration puts recursive self-improvement within reach by 2027 — birchlse · 2026-11-11
- "Retrofitting electricity to coal-powered factories": one line on AI's institutional mismatch — curious_vii · 2026-11-11