AI Could Replace 11.7% of US Jobs, MIT Study Reveals

The episode, an excerpt from the MIT Iceberg Index report, outlines the rapid shift in the economic viability of artificial intelligence within the U.S. labor market. The primary metric reveals that existing AI systems are already cost-competitive with median human wages for tasks representing 11.8% of total U.S. labor compensation, a finding that measures current economic replaceability rather than just theoretical capability. This transition is concentrated in white-collar knowledge work and is accelerating rapidly, having grown fourfold in only two and a half years due to factors like drastically falling model prices and the maturation of autonomous agent software. Several large corporations have already confirmed significant headcount reductions directly attributable to AI integration, leading to revised forecasts that predict millions of U.S. job displacements by 2028. Consequently, policymakers are grappling with proposed solutions, such as implementing an "AI Job Loss Dividend" or offering payroll tax holidays for new AI-adjacent roles, to mitigate the coming economic disruption. The report ultimately projects that the replaceable share of labor compensation could exceed 50% by the early 2030s, stressing the urgent need for strategies that buy time for human capital to reallocate.