Anthropic Economist: AI Hasn't Caused Unemployment Yet
Data-driven rebuttal to claims of AI wiping out white-collar jobs.
1Contradictory Predictions
Despite CEO Dario Amodei's warnings that AI could eliminate half of white-collar jobs soon, Anthropic economist Peter McCrory presented data showing no significant impact yet. US unemployment remains at 4.2%, considered full employment level.
2Timeline Discrepancy
While Amodei predicted massive job losses within 1-5 years in 2025, McCrory's 18-month internal research found no material effect currently. Unemployment in Claude-exposed jobs shows no relative increase compared to other occupations.
3Human-AI Collaboration
McCrory argues AI currently serves as 'thought partners' rather than replacements. Complex work still requires human oversight. Workers with domain expertise recover better from AI errors, contradicting simple substitution scenarios.
4Young Workers at Risk
Data shows hiring slowdown for young workers in AI-exposed roles. Technical writers, data entry and customer service jobs are projected to grow slower through 2034, acting as 'canaries in the coal mine'.
5Jevons Paradox Debate
Amodei's '90% automation leads to 10x productivity' theory aligns somewhat with McCrory's 'skill-biased' framework. But McCrory notes markets need adjustment time, while AI's rapid advancement may disrupt this balance.
6Future Uncertainty
McCrory acknowledges Amodei's 'singularity' scenario becomes possible if AI begins recursive self-improvement. However, he disputes the near-term timeline and scale predicted by his CEO, leaving Anthropic straddling two narratives.
Original article: Fortune | FORTUNE — Anthropic Ekonomisti: AI Şimdilik İşsizlik Yaratmadı