Scientific Sense ® Jon Hartley of Hoover institution on De-dollarization, AI, and tariffs.
Mar 1, 2026
Jon Hartley, economist and Hoover Institution policy fellow, discusses de-dollarization and why the dollar still dominates. He maps currency alignments and explains why reserve shifts are rare. He covers who uses generative AI at work, adoption limits, short-run labor effects and time savings. He analyzes tariffs, portfolio-driven dollar moves, and how policy uncertainty shapes growth.
AI Snips
Chapters
Transcript
Episode notes
Dollar Dominance Persisted Through Recent Geopolitical Shocks
- The US dollar remained dominant through late 2023 despite talk of de-dollarization after major shocks like COVID-19 and the 2022 Russia invasion of Ukraine.
- Central bank FX reserves (~59%), FX transaction pairs, trade invoicing (~50%), and global debt denomination still put the dollar far ahead of the next currency, with only modest historical variation between ~50–70%.
LLM Adoption Concentrated Among Young Skilled Workers
- Generative AI adoption surged early after ChatGPT's release then softened, with highest use among younger, higher-educated, higher-income workers in customer service, marketing, and IT.
- Usage varies by country: US and Anglosphere high, Italy Japan Canada low, and firms sometimes block public LLMs for legal and data-sensitivity reasons.
AI Gives Time Savings But Little Early Job Disruption
- Early labor-market effects of generative AI show small wage gains in more-exposed occupations but no measurable changes in job openings or total employment so far.
- Survey respondents report substantial time savings on tasks, implying productivity gains that may take time to flow into wages.