Professor Says AI Won’t Kill 50% of Entry-Level Jobs, Raising Value of Human Judgment
Updated
Updated · The Washington Post · Jul 11
Professor Says AI Won’t Kill 50% of Entry-Level Jobs, Raising Value of Human Judgment
1 articles · Updated · The Washington Post · Jul 11
Summary
Up to 50% job-loss warnings from AI executives overstate the threat, Victor Menaldo argues, because jobs are bundles of tasks tied together by judgment, coordination, trust and accountability.
More than 5% graduate unemployment and a 16% relative employment drop in AI-exposed entry-level fields show real pressure, but he says that reflects a painful transition rather than a white-collar extinction event.
AI’s core weakness is verification: it can generate fluent answers but cannot reliably audit its sources, show error bars or flag hallucinations, leaving humans to judge what is safe to use.
That shifts the premium to editors, doctors, lawyers and other workers who can catch errors, while firms that cut those roles risk quality-control failures and legal liability when bad output reaches decisions.
Menaldo likens AI to electricity or mechanized looms: a general-purpose technology that reorganizes work and rewards workers who master it, rather than simply subtracting jobs.