Anthropic researchers release economic scenarios on AI impact
anthropic working paper "economic scenarios for transformative ai" by five researchers presents three scenarios: modest 4% ai tasks, gdp 1.6% higher; substantial 12% tasks, gdp 8.3% higher; extreme 30% tasks, gdp 32.4% higher. the paper is not a prediction.
Key points
- Modest scenario: AI in 4% of tasks, GDP 1.6% higher, labor share 59.4%, total worker income +0.6%.
- Extreme scenario: AI in 30% of tasks, GDP 32.4% higher, labor share 45.2%, total worker income +0.5%, cognitive worker unemployment 17.9%.
- Headcount cuts precede wage cuts; wage flexibility can shift cognitive worker wages from 42.2% drop to 2.8% rise.
the model focuses on cognitive workers, 62.4% of us employment in 2025. it shows headcount cuts precede wage declines, with extreme scenario unemployment 17.9% for cognitive workers, overall 11.9%. wage flexibility changes can swing cognitive worker wages from 42.2% decrease to 2.8% increase.
labor share falls from 60.0% to 45.2% in extreme scenario, keeping total wages roughly unchanged. the report highlights that ai could lower wages for many while boosting capital returns, and that new task creation is limited in the extreme case.
[Deep Dive] Will Salaries Drop as AI Becomes More Powerful? ~Reading 3 Economic Scenarios from Anthropic Researchers Through 'Tasks'~
note.com · 2 October 2026
Loading the full article…
This text was published by note.com and written by おにぎりテック. It is reproduced here with attribution so you can read it in full; the rights remain with the publisher. Read it at the source ↗
The headline, key points and digest above were generated by Digest AI's editorial model from the linked sources. Automated summaries can contain errors: the sources are the record. Spotted a mistake? Tell us. Published by Martin K., who runs Digest AI and handles corrections.
More in Research
All →- Swarmchasers forum has 400 members, including Nightingale Collective and Transluce · 1 src
- UCLA builds optical AI that spots deepfakes with nearly 98% accuracy · 1 src
- Chalmers researchers build AI system to run yeast experiments · 1 src
- NVIDIA’s early support from Sega saved its GPU tech for AI · 1 src
- Researcher tests GPT-6 Astra in WoW with agent-wow framework · 1 src
Comments
via GitHub Discussions