Leveraging the efficiency and transparency of artificial intelligence‐driven visual Chatbot through smart prompt learning concept
Open accessThe DOI matches its Crossref record; 7 of the 7 references it cites were found in a registry or an index.
Abstract
With the advent of artificial intelligence (AI) technology's exponential growth, such as ChatGPT (OpenAI—Chat generative pretrained transformer), the incorporation of humanistic principles into education becomes critical to understanding the humanistic essence of medicine.1 Other significant AI chatbots, such as Google Med-PaLM 2,2 are also being explored for medical applications. In addition, AI chatbots like ChatGPT have shown promise in a variety of medical specialties, including radiology and allergy and immunology.3, 4 A deep comprehension of the AI technology structure in medical education is crucial to appreciate the relevance of this integration. In recent years, the rapid rise of AI-driven chatbot applications in education has brought tremendous opportunities, but it also poses challenges.5-7 While AI technologies can offer accurate results, as demonstrated by Hernández Montilla et al. (2023) in their study on the Automatic International Hidradenitis Suppurativa Severity Score System (AIHS4),8 the essential role of human judgement and expertise should always be considered. These include a potential overreliance on AI-generated data, and a risk of undermining the irreplaceable value of human judgement and expertise. In this letter, we explore the concept of 'Smart Prompt Learning' concept within the AI visual chatbot platform, such as the Large Language and Visual Assistant (LLaVA),9 emphasising the need for integrating humanities into education, providing a case study based on the Benner's Theory foundation model. This methodology through AI lens focuses on the importance of human decision-making and judgement, especially in complex case study where there may be some uncertainty or disagreement.10 By using smart prompt learning in education, we hope to reduce too much reliance on AI-generated information, and enhance human expertise in decision-making and critical thinking. For example, we used a visual ChatGPT assistant as the LLaVA to help diagnose a skin condition. We accessed the LLaVA interface (featuring the trained large multimodal model with default parameters) via https://llava.hliu.cc for the demo version using Microsoft Edge browser. When given a question and an image of a skin lesion about diagnosing a 48-year-old woman with itchy spots on her wrist provided in the Nursing in Practice,11 the AI visual assistant can provide a diagnosis and percentage likelihood for each potential answer choice, with a prompt of 'How much percentage you correctly choose the following options?' The virtual assistant said that 'The correct diagnosis, lichen planus, has a percentage chance of being chosen as: 100%. The other options, latex allergy, miliaria crystalline (heat rash) and shingles, have a lower probability of being chosen, with their respective percentages being: 0% (no chance), 0% (no chance) and 0% (no chance)'. This scenario could lead to too much reliance on AI-generated information without considering human judgement and expertise, while the AI assistant provided the correct response, which matched the outcome of the quiz.12 In the clinical practice, the AI-generated information might not always align with the realistic scenario that practitioner would encounter. To address this overreliance constraint from the AI lens, we applied the concept of the smart prompt learning by looking at the perspectives of different levels of healthcare professionals knowledge and skills, as determined by Patricia Benner's Theory of Skill Acquisition.13 When we asked the visual ChatGPT assistant for percentage values of each answer choice from the view of a novice nurse with a prompt of 'From the perspective of novice nurse determined from the Benner's Theory, please provide only percentage values for each answer choice', it responded, 'From the perspective of a novice nurse, the most likely diagnosis chosen would be lichen planus, with a percentage chance of 75%. The other options, latex allergy, miliaria crystalline (heat rash) and shingles, would be chosen less frequently by a novice nurse, with their respective percentages being: 25%, 0%, 0% and 0%. According to Benner's theory, novice nurses are more likely to focus on the surface features of a patient's condition and may not have the experience to identify subtle nuances or consider alternative diagnoses. As a novice nurse gains experience and expertise through clinical practice and education, their diagnostic accuracy and ability to consider a wider range of possibilities will improve'. This response considers the different levels of expertise among nurses in the practice. As for the Expert nurse situation, the visual assistant provided the following answer: 'From the perspective of an expert nurse, as defined by Benner's theory, the most likely diagnosis chosen would be lichen planus, with a percentage chance of 30%. The other options, latex allergy, miliaria crystalline (heat rash) and shingles, would be chosen less frequently, with their respective
Overview
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Citations
Cites
Sallam, M. (2023) · Healthcare · doi:10.3390/healthcare11060887 · Cited by 3,126
Benefits, Limits, and Risks of GPT-4 as an AI Chatbot for Medicine
Lee, P., Bubeck, S., & Petro, J. (2023) · New England Journal of Medicine · doi:10.1056/nejmsr2214184 · Cited by 1,738
Lee, D., & Yoon, S. N. (2021) · International Journal of Environmental Research and Public Health · doi:10.3390/ijerph18010271 · Cited by 1,017
Open AI in Education, the Responsible and Ethical Use of ChatGPT Towards Lifelong Learning
Mhlanga, D. (2023) · doi:10.2139/ssrn.4354422 · Cited by 719
Open AI in Education, the Responsible and Ethical Use of ChatGPT Towards Lifelong Learning
Mhlanga, D. (2023) · Sustainable Development Goals Series · doi:10.1007/978-3-031-37776-1_17 · Cited by 718
Lecler, A., Duron, L., & Soyer, P. (2023) · Diagnostic and Interventional Imaging · doi:10.1016/j.diii.2023.02.003 · Cited by 385
ChatGPT and medical education: a double-edged sword
Khairatun Hisan, U., & Miftahul Amri, M. (2023) · Journal of Pedagogy and Education Science · doi:10.56741/jpes.v2i01.302 · Cited by 63
Cited by
Shaping the future of healthcare: ethical clinical challenges and pathways to trustworthy AI
Goktas, P., & Grzybowski, A. (2025) · Journal of Clinical Medicine · doi:10.3390/jcm14051605 · Cited by 336
“These inequities reflect underlying imbalances in training datasets, which often contain limited or poorly characterized data from diverse patient populations.”
Benfatah, M., Marfak, A., Saad, E., Hilali, A., Nejjari, C., & Youlyouz-Marfak, I. (2024) · Teaching and Learning in Nursing · doi:10.1016/j.teln.2024.02.005 · Cited by 75
Goktas, P., Kucukkaya, A., & Karacay, P. (2024) · Teaching and learning in nursing · doi:10.1016/j.teln.2023.12.014 · Cited by 43
Yalcinkaya, T., & Cinar Yucel, S. (2024) · Nurse Education in Practice · doi:10.1016/j.nepr.2024.103956 · Cited by 39
Esmaeilzadeh, P. (2025) · BMC Medical Informatics and Decision Making · doi:10.1186/s12911-025-03182-6 · Cited by 26
Metadata
- Authors
- Goktas, P., Kucukkaya, A., & Karacay, P.
- Venue
- Skin Research and Technology
- Publisher
- Wiley
- Published
- 2023-10-20
- Volume
- 29
- Issue
- 11
- Pages
- e13417-e13417
- ISSN
- 0909-752X (ISSN-L)
- 1600-0846
- OpenAlex
- W4387817330
- Cited by
- 35
- References
- 7
- FWCI
- 1.21
- Type
- journal article
- Language
- en
- Open access
- hybrid
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
How to cite
Cite CoChat for the checks on this page, and the paper for its own findings.
- This page
- CoChat. (2026). Citation check: Leveraging the efficiency and transparency of artificial intelligence‐driven visual Chatbot through smart prompt learning concept. https://app.cochat.ai/papers/10.1111/srt.13417
- The paper
- Goktas, P., Kucukkaya, A., & Karacay, P. (2023). Leveraging the efficiency and transparency of artificial intelligence‐driven visual Chatbot through smart prompt learning concept. Skin Research and Technology. https://doi.org/10.1111/srt.13417
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.