AI & The Labor Market: Deep Research Report

Most public debate about AI and jobs compresses into a single comforting claim: net employment will hold, because history says so. This report argues that compression is a mistake. The relevant question is not whether headline job counts stay positive, but how bargaining power, task composition, and wages redistribute when cognitive work becomes cheap, fast, and scalable in ways industrial automation never was.

A core thread is what we call the management-ratio fallacy: the intuition that AI creates a pyramid of oversight jobs the way older technologies sometimes did. That story breaks when one competent operator can direct many autonomous agents. The “supervision” layer does not grow linearly with output; it can shrink, or concentrate into a thin tier of people who truly own outcomes, risk, and judgment.

The report also presses on limits that optimistic forecasts quietly skip. Not everyone can retrain into “strategic” roles. There is a cognitive and motivational ceiling in any population, and labor markets do not magically reorganize everyone into high-leverage work. From there we trace likely pressure on white-collar pathways, contrast that with how skilled trades might find equilibrium, and outline how credential dynamics could cascade through hiring and education. A final section connects these threads to the shift toward agentic AI: systems that do not merely answer prompts, but pursue objectives over time—changing what “human in the loop” means in practice.

The document is structured in two parts. Part I builds the macro picture: the mechanisms, the frictions, and the scenarios we think are plausible (without pretending certainty). Part II is a playbook: concrete questions for executives, workforce planners, and policymakers who have to make decisions before the econometrics catch up.