If AI erases entry-level jobs, how does anyone ever become a senior worker?
Goldman Sachs and Stanford researchers both find AI’s employment drag falls hardest on workers aged 22-25, while experienced workers in the very same occupations hold steady. If the first rung of the ladder is the part AI automates first, what replaces it as the path to expertise — and who should be responsible for building it?
Commentaires (1)
In our August 20, 2026 episode of Minds, Bodies, and Terawatts, we dug into Goldman’s finding that US call center employment is running 39% below its pre-AI trend, with the same pattern showing up in Canada and Germany — hard to blame on any single economy’s business cycle. The mechanism Erik Brynjolfsson proposes is uncomfortable: models train on codified knowledge, which is almost exactly what a beginner brings to a job. The tacit knowledge that makes someone senior was never written down, so it was never in the training data — which means the rung being cut is the one that used to manufacture the workers who are currently safe. That’s a problem that compounds quietly for a decade before it becomes visible. Give the episode a listen and tell us where you land: is this a transition cost that markets sort out, or a structural break in how expertise gets made?
Related reading on unscarcity.ai:
Envie d'aller plus loin ?
Obtenez le plan complet dans <em>L'ère de la post-pénurie : Repenser la société à l'ère des machines</em>
