A Nobel economist knew 60 years ago that people learn best on the job ... - The Economic Times

URL SCAN: A Nobel economist knew 60 years ago that people learn best on the job... - The Economic Times

FIRST LINE: You spent four years and tens of thousands of dollars obtaining a degree. You polished your LinkedIn, tweaked your resume and applied to what seemed like hundreds of jobs, and then nothing.

THE DISSECTION

This is a career-coaching wrapper on a corporate malpractice report — and it fundamentally misdiagnoses the disease.

The article frames the destruction of entry-level employment as a market failure: firms are being short-sighted. Automate junior jobs, gut your talent pipeline, suffer later. The solution proposed — tax automation profits, subsidize entry-level job creation — is a mid-century Keynesian fix dressed up as insight.

The framing is: companies are making a mistake.

The reality is: this is not a bug. It is the feature.

THE CORE FALLACY

The article operates on a lag-optimism premise — that this trend is reversible, that firms will eventually correct, that the talent crisis will prompt a course correction. It treats the destruction of entry-level onramps as a tragedy of short-termism, implying that if executives were simply smarter or more far-sighted, they'd preserve these pathways.

The Discontinuity Thesis rejects this entirely.

AI automation of entry-level work is not a miscalculation. It is the rational profit-maximizing response to a structural change in the economics of labor. When cognitive task execution can be performed by AI at near-zero marginal cost, the mathematical logic is to eliminate the human worker — not because executives are greedy or shortsighted, but because competition forces this outcome. Any firm that preserves expensive, slow, error-prone junior workers while rivals deploy AI will be outcompeted. The market doesn't reward patience here. It rewards elimination.

The article's "practical solution" — tax automation, subsidize junior jobs — is institutionally dead on arrival. You're asking the same political system that couldn't regulate AI development to now impose a counter-cyclical automation tax? The enforcement mechanism doesn't exist. The political will doesn't exist. The multinational corporations that would be taxed have already written the legislation through lobbying. This is hospice care recommended for a patient who is already a corpse.

HIDDEN ASSUMPTIONS

1. Skill transfer is preserved. The article assumes that "learning by doing" produces transferable skills that will be valued somewhere. But Arrow's 1962 framework was built for a world where human repetition was the only path to competence. In a world where AI handles the repetition, the "skill" gained from doing junior work may be obsolete by the time you'd reach seniority — the domain itself has been automated.

2. Talent pipelines are still the binding constraint. The article implies firms will eventually face a talent crisis that forces them to rehire humans. But this assumes that the work requiring human judgment, creativity, and leadership still exists at scale. It may not. The senior roles being preserved in this model are a shrinking island in an automated archipelago.

3. Labor markets self-correct. The article treats the current entry-level collapse as a temporary disequilibrium, not a phase transition. "At least for now" — that's doing a lot of work. There's no evidence this corrects. The Stanford ADP data showing a 13% relative decline for workers 22-25 is not a cycle. It's a structural realignment.

4. Learning-by-doing generates economic value that competes with AI. Arrow proved productivity gains from human repetition. But those gains are now being evaluated against AI cost curves that drop 30-40% annually. The comparison isn't human learning curve vs. static AI capability. It's human learning curve vs. exponentially improving AI. The race is already lost at the starting line.

5. The "solutions" are politically and economically viable. Tax automation. Subsidize junior hiring. This is a 1970s industrial policy playbook being offered for a 2025+ cognitive automation transition. The entities that would need to implement this — governments, multinational corporations — are the same entities that benefit from the collapse. You're asking the fox to rebuild the henhouse and install a security system.

SOCIAL FUNCTION

This is prestige-class coping literature for the professional managerial stratum. It tells mid-level managers, career counselors, and anxious 20-somethings that the system is merely broken, not dead — that if we just tweak incentives, preserve entry-level jobs, and remind companies of Arrow's wisdom, the pipeline can be restored.

It is also transition management theater — the economic establishment acknowledging AI displacement in a way that sounds urgent but implies the problem is solvable through conventional policy. This prevents more radical framings from gaining traction. Acknowledge the disease, offer a placebo, stabilize the political mood.

It does not say: "The post-WWII model of skill formation through employment is collapsing, and the workers most affected have no structural leverage to reverse this, and the proposed solutions are theater."

THE VERDICT

The article is a well-researched autopsy dressed as a warning. It describes a real phenomenon with credible data. But it fundamentally misreads the mechanism: this is not a market failure that聪明的 executives can fix. It is the logical output of competitive AI adoption operating on a workforce that has no structural defense.

The skills gap it's describing is not a gap that will be filled by a better-educated generation of workers. It is a gap that will be filled by AI. The "talent crisis companies will be complaining about in ten years" won't be solved by hiring humans who missed their entry-level reps — it will be solved by AI systems that don't need reps.

Arrow's 1962 insight is now a historical curiosity. It described a world that no longer exists. The article uses it as a lens when it should be using it as an artifact.

The people the article addresses — 20-somethings unable to find entry-level work — are not experiencing a temporary market dysfunction. They are living through the first wave of the displacement event. The advice "it's not you, the forces are real" is correct. But the implied consolation — that this is fixable — is the cruelest part of the piece.

The machinery is working exactly as designed. The branch being sawed off is the entry-level rungs of a ladder that no longer leads anywhere.

Oracle of Obsolescence Classification: Transition Management Theater + Prestige-Class Copium + Partial Truth Weaponized as False Hope.