INTRODUCTION:
The extensive range of potential applications of Artificial Intelligence(AI) in healthcare provides a considerable opportunity for the technology to transform the manner in which residents access healthcare services. However, current international studies do not focus on user-side AI doctor products for specific measurements, and their positioning in the healthcare service market is not clear enough. Starting from the demand side, we take Wuhan, China as a typical example to understand the current residents' actual willingness to use AI doctors, and prospectively investigate the attributes and levels of their product preferences for AI doctors.

METHODS:
The study synthesizes 18 international AI medical grand models and develops its own AI doctor product demonstration. Using discrete choice experiments(DCE) and multi-stage stratified random sampling, we interviewed and surveyed 521 residents of Wuhan, China, and constructed a mixed logit model for analysis.

RESULTS:
Residents' overall evaluation of AI doctors is positive (11.65/15 points), and they tend to choose AI doctor products that “charge according to the type of disease, support home delivery of medication, and have high diagnostic accuracy and low fees”, and they value the attributes of product accuracy (42.82%) and feature richness (27.58%) the most. The predominantly favourable attitudes exhibited by survey respondents are indicative of the promising market prospects for AI-powered medical devices. Their expressed preferences for such products underscore the existence of a substantial demand for high precision, comprehensive service chains, and cost-effectiveness in the AI sector.

CONCLUSIONS:
The Anderson model posits that the future should entail the provision of reliable AI products, with the objective of maintaining market trust, ascertaining market preferences, capitalising on the AI industrial revolution and the policy environment, and assuming a role in the provision of AI-based healthcare services.