Is AI coming for our jobs and wages? Past predictions offer some clues (www.rnz.co.nz)

A new meta-analysis reviewing 52 studies and 2,586 individual estimates finds little consistent evidence that robots and automation have systematically reduced wages. Prompted by early high-profile forecasts (Frey & Osborne’s 2013 claim that ~47% of U.S. jobs were at risk and Acemoglu & Restrepo’s 2017 evidence that robots depressed employment and wages), hundreds of follow-up studies produced mixed results: some show wage declines, others increases, but the pooled effect on wages is effectively zero and falls below conventional thresholds of economic significance. A separate University of Canterbury meta-analysis reached a similar conclusion for employment, suggesting the feared robo-revolution has not uniformly materialized. The findings matter because they shift the debate from alarmism to nuance. Automation’s impact is heterogeneous—routine cognitive and physical roles are more exposed, while creative and complementary jobs gain importance—so outcomes vary by industry, country, and worker group. The practical implications: policymakers should prioritize upskilling and facilitating human–AI collaboration rather than trying to halt automation, and employers should adapt business models and workforce development to capture productivity gains. In short, automation reshapes labor markets unevenly, creating pockets of displacement alongside new opportunities.