AI is not yet replacing workers in the US, researchers find

A recent Yale University study indicates that generative AI, exemplified by ChatGPT, has not yet caused significant disruption or job displacement in the US labor market since its 2022 launch, undercutting widespread fears of automation eroding demand for cognitive labor. While some companies have attributed recent layoffs to AI and executives foresee future impacts, the research highlights the early stage of AI adoption and the current absence of broad market upheaval, with researchers planning ongoing monitoring of its evolving effects.

Analysis

A recent Yale University study provides empirical evidence that generative AI has not yet induced a significant disruption in the broader U.S. labor market since ChatGPT's 2022 launch, directly countering the prevailing narrative of imminent, widespread job displacement. The analysis, which examined job distribution changes over 33 months, found no discernible impact on the demand for cognitive labor across the economy. However, this macro-level stability contrasts with micro-level actions and executive sentiment. Specific firms, including Dropbox (DBX) and Duolingo (DUOL), have explicitly cited AI as a rationale for recent layoffs, reflecting a strategic pivot towards automation. This is further supported by cautionary outlooks from industry leaders, such as the CEO of Salesforce (CRM). The practical limitations of current AI adoption are also becoming evident, with a Massachusetts Institute of Technology report indicating that 95% of companies experimenting with AI are not yet generating profits from it. Furthermore, the phenomenon of "workslop," where AI-generated output requires significant human correction, suggests that productivity gains may be offset by new hidden costs, complicating the investment case for AI-driven efficiency.

Key Decisions for Investors

• Investors should temper expectations for immediate, broad-based labor cost reductions from AI, as macroeconomic data from the Yale study indicates the disruption narrative is currently ahead of reality.
• Scrutinize company-specific claims of AI-driven efficiency, such as those from Dropbox and Duolingo; assess whether announced layoffs reflect genuine, sustainable margin improvement or mask deeper challenges in monetizing AI, as highlighted by the MIT report.
• Differentiate between long-term AI potential and near-term profitability by monitoring for tangible return on investment rather than just announcements of AI adoption, given the noted implementation hurdles and hidden costs like 'workslop'.