🔍AI Benchmark Map Covers 18,797 Tasks Across 923 Occupations
AI Benchmarks Skew Toward Big Jobs, Missing the Middle
TL;DR
A new AI benchmark map covers 18,797 tasks across 923 occupations but heavily focuses on large labor markets, leaving smaller, middle-income jobs underrepresented. 90% of American working hours are spent on tasks with little benchmark coverage.
The US economy has a new AI benchmark map that covers 18,797 tasks across 923 occupations. However, the map reveals a significant skew towards larger labor markets, with smaller, middle-income knowledge work jobs underrepresented. This means that while 90% of American working hours are spent on tasks with little or no coverage from existing benchmarks, the AI community is prioritizing the largest labor markets first. This focus on big jobs leaves many smaller roles without the necessary benchmarking to assess AI impact and potential automation.
Key Points
The benchmark map covers 18,797 tasks across 923 occupations.
Only 208 occupations have a task with at least 5 benchmark prompts.
6 of the 22 job families have no occupations covered by AI benchmarks.
90% of American working hours are spent on tasks with little benchmark coverage.
The AI community focuses on the largest labor markets first, undercovering smaller roles.
Why It Matters
If you're in a smaller, middle-income knowledge work job, the lack of AI benchmark coverage means your role's potential for automation and impact from AI is underassessed. For instance, AI-exposed lawyers' tasks are heavily covered, but legal secretaries and court clerks barely get any coverage. This disparity could lead to skewed economic assessments and missed opportunities for smaller roles to leverage AI.
Comments
Be the first to comment
Enjoyed this article?
Get it daily. 7am. Free. Reads in 5 minutes.