Related Experiment Video
Updated: May 13, 2026
Measuring Neural and Behavioral Activity During Ongoing Computerized Social Interactions: An Examination of Event-Related Brain Potentials
Published on: November 15, 2014
The Impact of ChatGPT Exposure on User Interactions With a Motivational Interviewing Chatbot: Quasi-Experimental
Jiading Zhu1, Alec Dong2, Cindy Wang1
1The Edward S. Rogers Sr. Department of Electrical & Computer Engineering, University of Toronto, Toronto, ON, Canada.
Background:
The worldwide introduction of ChatGPT in November 2022 may have changed how its users perceive and interact with other chatbots. This possibility may confound the comparison of responses to pre-ChatGPT and post-ChatGPT iterations of pre-existing chatbots, in turn affecting the direction of their evolution. Before the release of ChatGPT, we created a therapeutic chatbot, MIBot, whose goal is to use motivational interviewing to guide smokers toward making the decision to quit smoking. We were concerned that measurements going forward would not be comparable to those in the past, impacting the evaluation of future changes to the chatbot.
Objective:
The aim of the study is to explore changes in how users interact with MIBot after the release of ChatGPT and examine the relationship between these changes and users' familiarity with ChatGPT.
Methods:
We compared user interactions with MIBot prior to ChatGPT's release and 6 months after the release. Participants (N=143) were recruited through a web-based platform in November of 2022, prior to the release of ChatGPT, to converse with MIBot, in an experiment we refer to as MIBot (version 5.2). In May 2023, a set of (n=129) different participants were recruited to interact with the same version of MIBot and asked additional questions about their familiarity with ChatGPT, in the experiment called MIBot (version 5.2A). We used the Mann-Whitney U test to compare metrics between cohorts and Spearman rank correlation to assess relationships between familiarity with ChatGPT and other metrics within the MIBot (version 5.2A) cohort.
Results:
In total, 83(64.3%) participants in the MIBot (version 5.2A) cohort had used ChatGPT, with 66 (51.2%) using it on a regular basis. Satisfaction with MIBot was significantly lower in the post-ChatGPT cohort (U=11,331.0; P=.001), driven by a decrease in perceived empathy as measured by the Average Consultation and Relational Empathy Measure (U=10,838.0; P=.01). Familiarity with ChatGPT was positively correlated with average response length (ρ=0.181; P=.04) and change in perceived importance of quitting smoking (ρ=0.296; P<.001).
Conclusions:
The widespread reach of ChatGPT has changed how users interact with MIBot. Post-ChatGPT users are less satisfied with MIBot overall, particularly in terms of perceived empathy. However, users with greater familiarity with ChatGPT provide longer responses and demonstrated a greater increase in their perceived importance of quitting smoking after a session with MIBot. These findings suggest the need for chatbot developers to adapt to evolving user expectations in the era of advanced generative artificial intelligence.
