A recent study by political scientists Anil Menon of UC Merced and Baobao Zhang of Syracuse University found that Americans remain largely unconcerned about losing their jobs to artificial intelligence, even when told such changes could happen soon.
The research involved a survey of 2,440 U.S. adults who were presented with different scenarios predicting the arrival of “transformative AI” capable of replacing human workers. The projected timelines for this technology ranged from as early as 2026 to as late as 2060. Respondents read about the rapid development of generative models like Genesis, Claude, and ChatGPT before being asked about their views on automation risks.
According to the study, shorter timeframes made participants slightly more anxious about potential job losses due to automation but did not significantly change their predictions regarding when such losses would occur or increase support for government policies like retraining programs or universal basic income.
“These results suggest that Americans’ beliefs about automation risks are stubborn,” the authors said. “Even when told that human-level AI could arrive within just a few years, people don’t dramatically revise their expectations or demand new policies.”
Menon and Zhang argue that these findings challenge the idea that making technological threats seem more immediate will encourage public backing for regulations or safety nets. Their work draws on construal level theory, which examines how perceptions of time affect risk judgments. Participants who were told AI breakthroughs were imminent were not much more alarmed than those given distant projections.
The survey was conducted in March 2024 and divided respondents into groups receiving vignettes stating job-threatening AI would emerge in either 2026, 2030, or 2060; a control group received no timeline information. Each vignette described experts predicting advances in machine learning and robotics could impact professions ranging from software engineers to teachers and nurses.
After reading the scenarios, participants estimated when various jobs might be automated, expressed confidence in those estimates, reported worry about job loss, and indicated support for policy responses including restrictions on automation and increased funding for AI research.
Exposure to any timeline increased awareness of automation risks overall. However, only those presented with the 2060 scenario showed a significant rise in concern over job loss—possibly because they viewed this forecast as more credible than claims of near-term disruption.
The study appears in The Journal of Politics at a time when there is ongoing debate over how large language models may affect employment. While some tech leaders predict human-level AI within a decade, critics believe such forecasts exaggerate current capabilities.
Menon and Zhang note that while Americans remain cautious rather than panicked about these developments—a perspective useful for policymakers considering interventions—their design focused only on timeline cues without examining other psychological factors like beliefs about economic trade-offs or expert credibility. They also acknowledge that surveys cannot track changes in individual perceptions over time.
“The public’s expectations about automation appear remarkably stable,” they said. “Understanding why they are so resistant to change is crucial for anticipating how societies will navigate the labor disruptions of the AI era.”
