TaskNet: a third of tasks workers do are missing from O*NET, worth ~$140B a year in wages
Diyi_Yang · x · 2026-10-28
Mercor, Stanford, NBER and the ONET Resource Center released TaskNet, an open-source task catalog updating the US Department of Labor's ONET taxonomy for the AI era.
Key findings
- Workers in 6 occupations perform 79 tasks absent from ONET, worth about $140B a year in wages
- Per the data: 52% of software developers' tasks are missing (23 of 44), personal care aides 50%, ER physicians 45%, lawyers 31%, elementary teachers 23%, dishwashers 15%
- 5 tasks were newly created or reshaped by AI adoption, e.g. 'design prompts, retrieval logic, and context pipelines for AI applications' and 'review AI-generated legal content'
Method
TaskNet uses a scalable interviewer to elicit tasks directly from workers, plus their judgment on automation-relevant features like output-quality judgeability, so estimates of AI's impact on jobs built on incomplete taxonomies can be corrected.
Related event: Stanford and partners release TaskNet, an AI-era task catalog for O*NET(2 posts)→
More from AGI Musings
- Nathan Lambert: an effective org is worth multiples in compute for training frontier models — natolambert · 2026-10-28
- tszzl: verifiable ML subproblems still need expensive GPU experiments to get numbers — tszzl · 2026-10-28
- Dario: Those who called his GPT-2 fears 'pseudoscience' in 2019 now all apply to AI labs — irinarish · 2026-10-28
- Most Chinese AI discourse is about appliances and onions, not AGI — mattsheehan88 · 2026-10-28
- 60% of arXiv text is now AI-written, with fully AI-authored papers at 25% — wandedob · 2026-10-28
- Do Neural Networks Stumble Into Intelligence by Chance? A Reddit Deep-Dive on Architectural Determinism — Lost-Quiet-2021 · 2026-10-28