JMIR Publications

ABSTRACT

Background:

ChatGPT is a sophisticated large language model of artificial intelligence (AI) that was created by OpenAI and released to the public in November of 2022. It is characterized by generating human-like responses to natural language. Users can hold a conversation with the model where they input a prompt and it outputs a response. It has many applications including email-writing, solving math problems, grammar-checking, generating answers to complex questions and more. Other similar chat-based AI apps include Google Bard, Microsoft Bing ChatGPT, Socrative by Google, Hugging Chat, Snapchat AI, Perplexity AI, and YouChat and many others. All of these applications are like ChatGPT in terms of AI generating natural responses to prompts. Due to the rise in popularity of ChatGPT and other chat-based AI in medical education and medical specialty, more research must be conducted to better understand how it can be efficiently incorporated in medical education.

Objective:

This study aims to assess the knowledge, attitude, and practice of Alfaisal University medical students regarding the use of ChatGPT and other chat-based artificial intelligence applications in medical education.

Methods:

This was a cross-sectional study. A questionnaire was distributed by social media channels, from October 8th, 2023, through November 22nd, 2023, to Alfaisal medical students who were 18 years or older. Only current Alfaisal University medical students in years 1 through 6 of both genders (aged 18 and above) were targeted by the questionnaire. The study was approved by Alfaisal University Institutional Review Board. A Chi-square test was conducted to assess the relationships between gender and year of study and familiarity and reason of usage.

Results:

A total of 273 responses were received, 93 of which were males and 180 of which were females. In terms of preclinical and clinical years, there were 218 and 55 responses, respectively. Overall, males (97.9%) significantly showed more familiarity with Chat-GPT compared to females (90.09%), (p-value = 0.026). Males also used Google Bard and Microsoft Bing ChatGPT more than females, (p-value < 0.001). Additionally, students in clinical years of studies used ChatGPT significantly more for general writing when compared to preclinical students (p < 0.05). Findings with a p-value of higher than 0.05 showed that 46.4% thought using ChatGPT and other chat-based AI apps for coursework was ethical, while 29.4% were neutral, and 24.2% thought it was unethical.

Conclusions:

Familiarity and usage of Chat-GPT and other chat-based AI apps is common among the students of Alfaisal University. The uses of such apps differ between males and females and preclinical and clinical year students Clinical Trial: NA