Related Experiment Video
Updated: Oct 9, 2025

Augmenting Large Language Models via Vector Embeddings to Improve Domain-Specific Responsiveness
Published on: December 6, 2024
Is Artificial Intelligence Customer Service Satisfactory? Insights Based on Microblog Data and User Interviews.
Tengfei Zhao1, Jingjing Cui1, Jiayu Hu1
1Beijing Key Laboratory of Applied Experimental Psychology, National Demonstration Center for Experimental Psychology Education (Beijing Normal University), Faculty of Psychology, Beijing Normal University, Beijing, China.
Consumers are less satisfied with artificial intelligence (AI) customer service than human service, citing poor problem-solving and lack of human touch. AI voice and attitude influence positive feedback.
Area of Science:
- Consumer Behavior
- Artificial Intelligence
- Human-Computer Interaction
Background:
- Increasing adoption of AI in customer service for cost reduction and efficiency.
- Limited understanding of consumer satisfaction with AI service agents.
Purpose of the Study:
- To investigate consumer attitudes toward AI customer service in China.
- To develop theoretical models of consumer sentiment and evaluation of AI services.
Main Methods:
- Analysis of 17,673 Weibo data entries (Study 1).
- Conducting 33 in-depth interviews (Study 2).
- Construction of sentiment and evaluation models for AI services.
Main Results:
- Consumers exhibit a more negative attitude towards AI customer service compared to human service.
- Key complaints include poor problem-solving, slow responses, and lack of human touch.
- Positive feedback is contingent on AI's voice traits and service attitude; AI is perceived as human-like and stress-relieving.
Conclusions:
- Chinese consumers generally hold negative attitudes towards AI customer service.
- AI service development should focus on improving problem-solving, response times, and incorporating human-like qualities.
- Findings offer actionable insights for enhancing AI customer service solutions.
Related Concept Videos
Non-equilibrium in the Cell
Stereotype Content Model
Impression Management Techniques II: Ingratiation
Data Collection by Survey
Impression Management Techniques III: Aligning Actions
Self-Evaluation Maintenance Model

