Collapse of social mobility

This risk is the closing of the entry rung projected forward in time. It is not that there are fewer jobs; it is that the path by which someone went from trainee to expert —doing for years the narrow work nobody else wants— stops existing as a route.

The starting point is measured: the youth employment divergence in exposed occupations reaches 19% in relative terms by June 2026 and works through less hiring ^[164]Canaries in the Coal Mine? Six Facts about the Recent Employment Effects of Artificial IntelligenceBrynjolfsson, Erik; Chandar, Bharat; Chen, Ruyu · 2026 · preprintView source ↗Accessed on 9 September 2026; the most specific cut is software developers aged 22 to 25, with a fall of almost 20% since 2024 ^[34]The AI Index 2026 Annual Report — Chapter 4: EconomyStanford Institute for Human-Centered AI · 2026 · reportView source ↗Accessed on 9 September 2026. The mechanism that connects it to mobility comes from the controlled experiments, and it is counter-intuitive: assistance benefits most those who start lowest —14% more issues resolved per hour on average and 34% among novices, in a deployment covering 5,179 support agents— ^[163]Generative AI at WorkBrynjolfsson, Erik; Li, Danielle; Raymond, Lindsey · 2025 · paperView source ↗verified through CrossrefAccessed on 9 September 2026. What the tool spreads around is the differential that made a trainee worth hiring. Drago and Laine name the shape it would take: pyramid replacement, with automation entering at the base of the org chart and moving up ^[336]The Intelligence CurseDrago, Luke; Laine, Rudolf · 2025 · reportView source ↗Accessed on 9 September 2026.

What this does not demonstrate. No intergenerational mobility series covers this period, let alone with a design that would allow a change to be attributed to AI. The alternative explanation still stands: Iscenko and Curto Millet show that youth fragility in slowdowns is nothing new and that job postings in exposed occupations also fell more in 2020 ^[566]Looking for the Ladder: Is AI Impacting Entry-Level Jobs?Iscenko, Zanna; Curto Millet, Fabien · 2026 · reportView source ↗Accessed on 9 September 2026. And there is one finding that cuts against the clean-substitution reading: outside the capability frontier, AI assistance reduced the correctness of 758 consultants by 19 percentage points, and the frontier is jagged —tasks of apparently similar difficulty fall on opposite sides of it and there is no telling which without trying— ^[327]Navigating the Jagged Technological Frontier: Field Experimental Evidence of the Effects of Artificial Intelligence on Knowledge Worker Productivity and QualityDell'Acqua, Fabrizio; McFowland III, Edward; Mollick, Ethan R. et al. · 2026 · paperView source ↗verified through CrossrefAccessed on 9 September 2026.