For centuries, rationality, language, and logic have been considered defining features of humanity. However, recent advances in machine learning applied to Large Language Models, such as ChatGPT, have made artificial agents increasingly human-like on these dimensions. How are people reacting to these resemblances between humans and AI? 

While some people see these developments as signs of ever-increasing productivity, enabling lives of leisure served by artificial agents, others fear dystopian nightmares of mass unemployment—or even direct confrontation with hostile AI beings. 

Our recent studies, conducted at Stanford University, show a more subtle yet fundamental change underfoot: The emergence of AI may already be shifting people's perception of what it means to be human—a phenomenon we call the "AI effect." These findings have significant implications for business, work, and life.

Demonstrating the "AI Effect"

In our research, we first categorized human attributes into two types: those "shared" with AI (such as logic, memory, and language) and those "distinct" from AI (such as humor, having a personality, and having beliefs)—and conducted a survey to make sure people agreed with our distinction.

Then we conducted four experimental studies with nearly 5,000 participants. Some participants read an article about advances in artificial intelligence (thus putting AI top of mind), while others read an article about other topics, such as trees, or read no article at all. Participants then rated a combination of traits as more or less essential to being human.

We found that the participants who read an article about advances in artificial intelligence later rated "distinct" human attributes as more important to being human. Simply reading about AI was enough to make respondents boost the relative importance of beliefs, desires, and relationships as more essential to being human. Interestingly, participants did not view the traits shared with AI as less essential to being human. They only boosted the set of traits capturing aspects of the human experience that people assume AI cannot yet emulate.

We call this the "AI Effect." 

What Are the Implications of the "AI Effect?" 

Just as the emergence of hand-held calculators made mental algebra irrelevant to demonstrating true intelligence, AI is likely changing what characteristics are considered essential to being truly human. In study after study, we observed that people exposed to AI emphasized various traits distinct to humans, such as having beliefs and having a personality, as core to the human experience. 

Our research suggests that, as AI becomes increasingly sophisticated, people are likely to place increasing importance on those aspects of their lives they think are still out of reach for AI.  They may even seek out more opportunities for human connection and meaningful spiritual experiences, which they view as distinctly human capacities.

What are the consequences of these changing perceptions of what it means to be human? Will organizations place greater importance on soft skills that help people stand out, express their values, and show up with their whole selves as AI becomes more ubiquitous?  Will students, workers, or consumers seek out more opportunities to express themselves, embrace their values, and connect with other humans over practicing skills best left to machines? 

Finally, the participants in our studies read about AI advances in the months before ChatGPT brought Large Language Models to mainstream consciousness. Do those who now routinely use these tools no longer feel threatened by AI advances?   


As AI continues to advance, more and more of the "distinctly human" traits are becoming increasingly "shared" with what artificial intelligence can do. If this rate of technological progress continues, what will people think it means to be human a few decades from now? Will they become inured to AI advances, and no longer feel the need to change how they think about being human? Or will they continue to find ways that distinguish humans from machines? Eventually, we may be forced to become more comfortable having nearly as much in common with our digital cousins as we do with one another.

For Further Reading

Santoro, E., & Monin, B. (2023). The AI Effect: People rate distinctively human attributes as more essential to being human after learning about artificial intelligence advances. Journal of Experimental Social Psychology107, 104464.

Erik Santoro is a Postdoctoral Scholar at the Columbia Business School and recently completed his PhD in Psychology at Stanford University.

Benoît Monin is the Bowen H. and Janice Arthur McCoy Professor of Ethics, Psychology, and Leadership, with joint appointments in the Stanford Graduate School of Business and the Department of Psychology.