Research shows human conversations offer value beyond chatbots

Associate Professor Jingbo Meng has co-authored a paper titled, "Examining the Content and Form of Supportive Conversations with Chatbots," recently accepted for publication in the International Journal of Human-Computer Interaction. Her research revealed that people were more open to discussing their problems in detail when talking to humans rather than chatbots.
Meng’s study explored interactions between individuals and either a chatbot or a human support provider. Meng analyzed 1,232 statements from 114 conversations finding five types of conversation patterns.
In addition to showing a preference for sharing problems with humans, Meng’s research proved that the key distinction between the various types of chatbots was that individuals are less likely to explain the problem to the expert chatbot than the novice one.
In an interview related to her work, Meng speaks on the possible advantages and disadvantages of AI companion bots for teenagers. She discusses the notion that these chatbots could offer emotional support and stimulate conversation. However, there is risk involved in terms of their impact on social development and mental health. The interview highlights the complexities of implementing AI in delicate situations like adolescent well-being and underscores how critical it is to understand the nuances of human-chatbot interactions.
Meng's recent work investigates the use of chatbots in online social networks to deliver social support and tackle health misinformation. Overall, her research focuses on how social networks and communication technologies impact physical and mental health. She strives to understand and create online social networks that enable social support, sharing of accurate health information, and influence health behavior change to improve psychological well-being.
Written by student Ellie Kulberg