as a form of communication is crucial. The study, conducted by researchers from Stanford University, the University of California, Berkeley, and the University of California, Santa Barbara, suggests that while AI-generated responses can be effective in making people feel heard, there is an underlying bias against AI that diminishes its perceived effectiveness.
The findings of the study highlight the complex relationship between humans and AI, and the need for further research and development to address the biases and limitations associated with AI communication. The study also emphasizes the importance of considering the psychological and emotional effects of AI-generated messages on human recipients, and the ethical implications of relying on AI for communication in various contexts.
The importance of addressing human psychological needs is becoming more relevant as empathetic connections in today’s fast-paced world continue to decline. Many individuals are experiencing unmet needs for being heard and validated.
Yidan Yin, Nan Jia, and Cheryl J. Wakslak from the USC Marshall School of Business conducted research to explore whether AI, which lacks human consciousness and emotional experience, can effectively make people feel heard and understood.
“In the context of a growing loneliness epidemic, a significant aspect of our motivation was to determine if AI can truly assist people in feeling heard,” stated the paper’s authors.The primary author, Yidan Yin, is a postdoctoral researcher at the Lloyd Greif Center for Entrepreneurial Studies at USC Marshall.
The research team’s discoveries not only emphasize the possibility of AI to enhance human ability to comprehend and communicate, but also raise fundamental questions about the significance of being understood and practical inquiries about the most effective ways to capitalize on AI’s strengths to promote greater human well-being.
Through an experiment and subsequent follow-up study, “we found that while AI shows greater potential compared to non-trained human responders in offering emotional support, the devaluation of AI responses poses a challenge.Nan Jia, an associate professor of strategic management, stated that effectively deploying AI’s capabilities poses a significant challenge. A research team at USC Marshall studied people’s emotions and perceptions after receiving a response from either AI or a human. The survey involved messages that were actually generated by AI or a human, and participants were informed of the true source of the message. The team discovered that regardless of the actual source of the message, people’s feelings and perceptions were influenced by whether they believed the message was generated by AI or a human.Cheryl Wakslak, associate professor of management and organization at USC Marshall, mentioned that the content and the perceived sender of the message were influential. She stated that people felt more acknowledged when they received a message from AI rather than a human, but if they thought the message came from AI, it made them feel less acknowledged.
Yin, in the study, found a bias against AI. Although AI is helpful, people generally do not like it.
Wakslak added that perceptions about AI are likely to change over time. She also noted that the two effects observed in their study were quite interesting.The impact of receiving a message from AI is quite similar in size to the response bias when the message is recognized as originating from AI. As a result, these two effects essentially cancel each other out. Participants also experienced an “uncanny valley” response, feeling a sense of discomfort when they learned that the empathetic response came from AI. This highlights the emotionally complex nature of interactions between AI and humans. Additionally, the survey asked participants about their overall willingness to engage with AI, which helped to moderate some of the effects, according to Wakslak.
“Individuals who have a more positive attitude towards AI do not experience the same negative response, and this is interesting because it raises the question of whether people’s attitudes towards AI will become more positive over time,” she said. “This remains to be seen but it will be fascinating to observe how this unfolds as people become more familiar with and experience AI.”
AI provides better emotional support
The research brought attention to important details. The reactions triggered by AI were linked to increased optimism and reduced distress, indicating a positive emotional impact on recipients. AI also displayed a more methodical approach than humans in terms of
The research found that AI was able to provide emotional support without overwhelming the individual with practical suggestions. Yin noted that it is ironic that AI was better at using empathetic and validating strategies, which have been shown to be effective in prior research. This suggests that humans may be able to learn from AI in providing validation when others are complaining about something.
Instead of replacing humans, the research emphasizes the different advantages of both AI and human responses. The advanced technology has the potential to empower humans to utilize AI as a valuable tool in providing support.
Assist people in understanding each other and learning how to respond with emotional support and validation.
The study’s results have significant implications for AI’s integration into social settings. AI’s abilities could offer a cost-effective and scalable solution for social support, particularly for those who are unable to access individuals who can provide this support. However, the research team emphasizes the importance of carefully considering how AI is presented and perceived in order to maximize its advantages.
