New paper uses 288M job postings to measure 'bottleneck pressure' where AI raises work value
soumitrashukla9 · x · 2026-10-22
A new paper, "Technology Bottlenecks: Measuring Where AI Could Raise the Value of Work It Cannot Yet Do," flips the usual AI-exposure framing: instead of asking what AI can already do, it measures where the marginal value of work AI cannot yet do may rise as AI speeds up the rest of a job.
Method: 288 million US job postings mapped line by line to their closest ONET tasks, combined with GPT-4 feasibility ratings from Eloundou et al. (2024), yield a task-level "bottleneck pressure" score for each task in each occupation. Commenter Adam Hall calls bottleneck pressure an important concept for understanding and measuring AI progress and its economic impact.
More from AGI Musings
- Dramstead quips that everything EA touches becomes corrupt — both EAs — mjdramstead · 2026-10-22
- Robots will strip human hours out of prices, leaving materials plus energy — Dr_Singularity · 2026-10-22
- Dan Faggella: bullish on Nvidia because its hardware may host souls — danfaggella · 2026-10-22
- As software turns probabilistic, interfaces shift from defining features to narrowing possibilities — round · 2026-10-22
- Former FTC Chair Lina Khan calls today's AI exceptionalism "ludicrous" in AI governance talk — AINowInstitute · 2026-10-22
- Constitutions for the multiagent era: how AI constitutions can govern agent organizations — sethlazar · 2026-10-22