Human Capital Repricing in the AI Economy

Gary Becker's 1964 human capital theory held that education and training are investments increasing productive capacity, rewarded by lifetime earnings premiums. The framework assumed that acquired skills remain scarce enough to command returns for a career's duration. AI tests this assumption empirically: returns to execution skills (coding, legal research, financial modeling) are declining wherever AI performs competently, while returns to judgment, taste, and integrative capacity—harder to teach, slower to develop—are rising. The repricing is not total erasure but violent reallocation. A computer science degree's value proposition shifts from certifying coding ability to signaling general cognitive quality plus network access. Universities built to teach execution face an existential crisis: their core product is being commoditized while the skills the market now prizes—judgment under uncertainty, cross-domain integration, question formulation—resist standardized instruction.