Chapter 17College, Careers, and the Rise of AI
—Morgan R. Frank, University of Pittsburgh
This chapter examines how large language models (LLMs)—such as ChatGPT, Claude, Copilot, and Gemini—are reshaping the labor market for college-educated workers, with an emphasis on early impacts since ChatGPT's November 2022 launch. Historically, higher education provided a buffer against automation, with white-collar roles benefiting from technology that enhanced productivity. LLMs, however, can now perform many cognitive tasks previously reserved for educated professionals, raising concerns about the stability and value of these jobs. Using data from millions of U.S. résumés compiled by Revelio Labs, the study investigates changes in salaries and job-seeking behavior among both recent graduates and more experienced workers, focusing on how these trends vary by an occupation's exposure to LLM automation risk.
The analysis finds an overall 8.6 percent decline in real salaries for jobs started after ChatGPT's release, with sharper drops in occupations highly exposed to LLMs. Before ChatGPT, recent graduates in high-exposure jobs enjoyed a 28.8 percent wage premium over peers in low-exposure jobs; this premium largely disappeared afterward. Recent graduates saw a 15.1 percent salary decrease in exposed roles, compared with a 6.6 percent decrease for non-recent graduates. Higher degrees did not fully shield workers—bachelor's and master's degree recipients faced the steepest declines, while ...
Become an O’Reilly member and get unlimited access to this title plus top books and audiobooks from O’Reilly and nearly 200 top publishers, thousands of courses curated by job role, 150+ live events each month,
and much more.
Read now
Unlock full access