Economist Podcasts Bot the difference: AI’s absence in economic data
93 snips
Feb 27, 2026 
Jon Fasman, senior culture correspondent who tells a portrait of Maine lobsterman Virginia Oliver. Ọrẹ Ogunbiyi, Africa correspondent reporting on brutal attacks and shifting militant dynamics in Nigeria. Alex Domash, economics correspondent exploring why AI has not yet shown up in macroeconomic data. They discuss AI’s measurable adoption and productivity puzzles, widening security crises in Nigeria, and a life spent at sea.
AI Snips
Chapters
Transcript
Episode notes
GDP Growth Outpaced Hiring In 2025
- Recent US data showed a puzzling gap: strong real GDP growth alongside weak employment growth.
- In 2025 real GDP grew ~2.2% while employment rose only ~0.1%, which would normally imply a surge in productivity but other factors complicate that story.
AI Investment And Labor Shortages Skew Productivity Measures
- Part of the measured output boost came from heavy AI infrastructure spending by firms.
- Simultaneously tighter immigration and falling temporary-worker participation removed many lower-productivity workers, biasing productivity measures upward.
Three Inputs Needed To Quantify AI's Productivity
- Measuring AI's impact requires three inputs: adoption rate, intensity of use, and per-use productivity gains.
- Studies converge: ~40% adoption, only ~13% use daily, averaging ~2 hours/week of AI use across workers.