Evaluating the Use of Large Language Models vs. Traditional Textbooks in Physiology Education: Insights into Generation Z Medical Students' Preferences

Background: Medical education is rapidly evolving with the integration of artificial intelligence [AI] and digital technology. Large language models [LLMs] such as ChatGPT, Bing, Bard, Gemini, Claude, and Perplexity are increasingly used by medical students primarily for their accessibility and ease of use. However, the abundance of information generated by these tools may be overwhelming and can complicate the independent assimilation of knowledge. Objectives: This experimental study aimed to evaluate the effectiveness and acceptability of LLM based AI tools compared with traditional textbooks for learning Physiology among Generation Z medical students. Methods: First-year MBBS students were randomized in to 2 groups: 1- standard Physiology textbooks and 2- LLM based AI tools. After one hour of self directed study, students completed a 15 item MCQs [topic specific] via Google form. The Medical Artificial Intelligence Readiness Scale for Medical Students [MAIRS-MS] was used to evaluate AI readiness and perceptions towards AI-assisted learning. Results: There was no significant difference in test scores between the standard Physiology textbook group and LLM based AI tools group across various Physiology topics. However, students responses regarding the Conclusion: Despite the growing interest and potential advantages of LLM based AI tools, this study found no significant preference or performance advantage for LLMs over traditional textbooks among Generation Z medical students to study Physiology. However, feedback responses from students preferred a combined use of both LLMs and textbooks rather than relying exclusively on a single learning resource, highlighting a shift in learning preferences among Generation Z medical students. Future research is warranted to explore the role of LLM based AI tools roles in medical education, with particular attention to global content validation, learning retention and ethical consideration.