Anthropic, Altman Recast AI Job Threat as 33% Task Coverage Falls Short
Updated
Updated · The Guardian · Jul 25
Anthropic, Altman Recast AI Job Threat as 33% Task Coverage Falls Short
3 articles · Updated · The Guardian · Jul 25
Summary
Anthropic’s March analysis found no systematic rise in unemployment for AI-exposed workers since late 2022, undercutting earlier warnings that AI could erase half of entry-level jobs within one to five years.
33% of computer-and-math tasks are currently covered by Claude, far below the near-100% theoretical ceiling, while labor-productivity gains have lagged the early IT boom despite surging datacenter spending.
Sam Altman said in May he no longer expects a jobs “apocalypse,” and economists including David Autor and Daron Acemoglu argue task-level automation has so far produced only modest overall employment effects.
The softer view comes as AI faces broader doubts over technical limits, weak economics and politics: the Nasdaq is down about 8% from its early-June peak, 70% of Americans oppose local AI datacenters, and datacenter power demand could top 945 terawatt-hours by 2030.