Benedict Evans argues that AI job exposure models may be less predictive than they look. Using accounting automation, CPA demand, newspapers, and Uber as examples, he shows how technology often changes businesses and job definitions in ways neat forecasts miss.
Predicting AI job exposure
Benedict Evans
Article
8 Pages
Key Takeaways
Automation Can Expand: Accountants rose from roughly 0.1% to 1.3% of US employment over the 20th century despite decades of software automation.
CPA Hiring Persisted: CPA firms hired around 28,000 accounting graduates in 2020, even after 50 years of financial automation reshaped workflows across firms nationwide.
Models Miss Context: Taxi disruption was hard to spot from 2005 smartphone exposure models, despite $1m medallion mortgages later becoming a major issue.
Join our newsletter to have all of this content + Exclusive Newsletter Bonus Content delivered to your inbox every week
