A new working paper by researchers at Stanford University and the Barcelona School of Economics argues that generative AI raises average wages by 21% and reduces wage inequality, chiefly through “simplification” that lowers skill barriers and lets less-experienced workers compete for tasks once reserved for specialists. The model predicts broad welfare gains at labor-market entry but significant job reshuffling: administrative roles contract while science occupations expand, and some high-skill categories—including architects, engineers and executives—see absolute wage declines. The authors say their predictions align with recent labor data, though impacts vary by occupation. White House AI and cryptocurrency czar David Sacks called the results a “narrative violation,” highlighting how the findings cut against common fears of AI-driven inequality. As a working paper, the results may evolve with further scrutiny and real-world evidence.
Related articles:
Generative AI at Work: Evidence from a Call Center
Automation and artificial intelligence: How machines affect people and places
