A multi-decade MIT study analyzing 35,000 US Census Bureau job categories and US patent text found that from 1940–2018, technology displaced jobs at an accelerating rate. Automation reduced employment at roughly twice the pace in the 1980–2018 period versus 1940–1980, while augmentation (technology creating new tasks) grew modestly and insufficiently to offset losses. The research confirms that net job creation from technology has lagged automation since 1980, with new work concentrating in high-skill and low-wage service roles, leaving a middle-class gap. For financial institutions, this provides quantified evidence of structural labor-market polarization driven by technology adoption, raising questions about future demand for AI-related services in a bifurcated labor market.
130 other tracked developments touching the same claims.