For a long time, the general public has regarded the release of ChatGPT in November 2022 as a "turning point" for the job market. However, a recent study jointly released by several top U.S. universities presents an unexpected conclusion: the stagnation of employment growth and the rising risk of unemployment in jobs most vulnerable to AI (such as computer science and mathematics) had already begun several months before the release of ChatGPT.

The research team conducted a deep analysis of 10.6 million LinkedIn profiles, 3 million university course syllabi, and data from the U.S. Department of Labor, finding that the risk of unemployment in the fields of computer science and mathematics had already begun to rise at the beginning of 2022. Surprisingly, after the official release of ChatGPT, the rate of increase in this risk actually slowed down, rather than accelerating sharply as people had feared.

This data discrepancy suggests that macroeconomic factors—such as the Federal Reserve's interest rate hikes and the demand adjustment for software development positions after the outbreak of the pandemic—may have a deeper impact on the job market than the spread of AI technology. Simply attributing the rise in unemployment to artificial intelligence may overlook the complex economic dynamics behind it.

Of concern to professionals is that, although the barriers to employment have increased, "AI skills" remain a key to high-paying jobs. The study shows that graduates who possess skills such as writing and programming—skills easily replaced by AI—actually received higher starting salaries and were hired faster after the release of ChatGPT. This indicates that companies are currently more eager for versatile talent who can skillfully use AI tools and verify AI outputs, rather than simply replacing human tasks with technology.