Entry-level jobs were actually secret apprenticeships all along, and AI just cut the subsidy

The labor market has tightened over the past decade, leading to a decline in entry-level jobs for those in occupations most exposed to AI. This decline began in March 2022, when the Federal Reserve started raising interest rates. Despite widespread panic that AI is destroying entry-level white-collar work and making college degrees obsolete, the evidence suggests otherwise.

A study by Stanford's Digital Economy Lab shows that workers aged 22 to 25 in AI-exposed occupations are 19% behind their peers in less-exposed fields, but this gap has been widening over the past year. However, this decline in job postings can be attributed to the tightening labor market cycle and the sensitivity of AI-exposed occupations to interest rates.

Young workers without college degrees also experienced rising unemployment at a similar pace. While Stanford researchers caution that these are descriptive patterns and not causal estimates, they have pushed back on the monetary policy story, noting that the most exposed jobs are not necessarily the most rate-sensitive, and the employment gap for young workers in exposed occupations continues to widen even as interest rates have come down.

The truth is that it is too early to know the exact impact of AI on the labor market, as it is impossible to separate the impact of overlapping shocks in real time. However, it is clear that firms are not firing their junior employees; instead, they are hiring fewer of them, and this decline in hiring is concentrating in sectors where AI is automating work rather than where it is complementing it.

Employers still want experienced people but have stopped funding the process that produces experience, viewing junior hiring as a cost line that automation has erased. Instead, they must recognize that it was never an operating expense but a capital investment mislabeled. This means rethinking the training offered, job rotations, and mentoring and coaching, all aimed at developing the judgment of tomorrow's senior professionals.

Universities face a similar problem, as the Ph.D. is an apprenticeship funded by the productive value of the apprentice's work. The editors of Nature have warned that early-career researchers now face the danger that tasks crucial to their training as scientists are done by a machine. Just as getting the right answer is not the point in forming judgment, the outcome of AI not being inevitable is dependent on employers continuing to fund the formation process, rather than treating it as a cost that can be cut.