A study published by the American Psychological Association found that employees who frequently interacted with artificial intelligence systems were more likely to experience loneliness, which could lead to increased insomnia and post-work drinking.
The researchers conducted four experiments in the United States, Taiwan, Indonesia and Malaysia. Results were consistent across cultures. This study was published online in the Journal of Applied Psychology.
Principal Investigator Dr. Pok Man Tang used AI systems in his previous job at an investment bank, which made him interested in researching timely issues.
“Rapid advances in AI systems are sparking a new industrial revolution, reshaping the workplace with many benefits, but also bringing with them some unknown dangers, including psychological and physical harm to employees. ,” said Tang, assistant professor of business administration at New York University. Georgia. “Humans are social animals, and separating jobs with AI systems can have a negative impact on employees’ personal lives.”
At the same time, working with AI systems may have some advantages. The researchers found that employees who frequently used AI systems were more likely to offer help to their colleagues, but that response may have been driven by feelings of loneliness and need for social contact. bottom.
In addition, the study found that participants with high levels of attachment anxiety (a tendency to feel insecure and worried about social connections) responded positively to working with AI systems, such as helping others, and Both negative responses, such as loneliness, were found to react more strongly. and insomnia.
In one experiment, 166 engineers at a biomedical company in Taiwan working on AI systems were surveyed over three weeks about loneliness, attachment insecurity, and belonging. Colleagues rated individual participants for beneficial behavior, and family members reported participants’ insomnia and post-work alcohol consumption. Employees who interacted more frequently with AI systems were more likely to experience loneliness, insomnia, and increased alcohol consumption after work, but also exhibited some helping behaviors toward fellow employees. I was.
In another experiment of 126 real estate consultants from a property management company in Indonesia, half were told not to use the AI system for three consecutive days, while the other half were told to use the AI system as much as possible. was done. Results for the latter group were similar to the previous experiment, except that there was no association between the frequency of AI use and post-work alcohol consumption.
Similar results were found in an online study of 214 full-time adults in the United States and an online study of 294 employees of a Malaysian technology company.
The findings are correlated, and it has not been proven that working with AI systems causes loneliness or any other reaction, there is simply an association between them.
In the future, Tan said, AI technology developers should consider equipping AI systems with social features such as human voice to emulate human-like interactions. Employers can also limit the frequency of work with AI systems and provide social opportunities for employees.
AI systems can focus on tedious and repetitive tasks, Tan added, while humans can do team decision-making and other tasks where social connections are important.
“Mindfulness programs and other positive interventions can also help reduce feelings of loneliness,” Tan says. “As AI continues to expand, we need to act now to mitigate potential negative impacts on those who use these systems.”
Articles: “Nobody’s an Island: Interacting with Artificial Intelligence to Uncover Work and Post-Work Outcomes,” Pok Man Tan, Ph.D., University of Georgia; Joel Koopman, Ph.D., Texas A&M University. Dr. Michael Mai Ke and Dr. David De Cremer, National University of Singapore. Dr. Jack H. Chan, Nanyang Business School, Ph.D. Dr. Philip Reinders, Cardiff University. Master Ching Tung Stewart Ng, Dr. Yi Heng Cheng, National Sun Yat-sen University. Journal of Applied Psychology, published online June 12, 2023.
