Authors

  1. Shay, Amy PhD, RN, APRN-CNS, FCNS

Article Content

The scientific discipline of artificial intelligence (AI) has evolved so rapidly in recent years that it prompted introduction of the 2020 National Artificial Intelligence Initiative Act by the US Congress to study its current and future impact. Although there appears to be no singular definition for AI, the US Department of State defines it as "a machine-based system that can, for a given set of human-defined objectives, make predictions, recommendations or decisions influencing real or virtual environments."1 Artificial intelligence has been deployed recently within the healthcare and education domains, making it imperative for nursing educators to understand its potential applications and limitations.2,3

 

Chat Generative Pretrained Transformer (ChatGPT) is an AI large language model developed by OpenAI4 and released to the public in November 2022. ChatGPT uses natural language processing through neural networks to generate human-like text responses to user queries.3,5 ChatGPT has demonstrated its power by passing law school examinations, the US Medical Licensing Examination, and producing full H&P (history and physical) notes and SOAP (subjective, objective, assessment, and plan) notes, helping make it the fastest-growing user application in history.2,3 These capabilities, however, have implications for both students and educators.

 

For all its power, ChatGPT is not without limitations. ChatGPT can provide articulate answers that appear reasonable but may be factually incorrect with false citations. The system may also provide false answers that are irrelevant or nonsensical. This phenomenon has been termed a "hallucination" and is blamed on weaknesses in the training data.2,3,5 Accuracy issues may also be compounded by the system's insufficient context to comprehend nuance and its reliance on data prior to 2022.2 Bias error may result from skewed training data leading to faulty predictions and potentially offensive text responses.3,5 Despite these limitations, the content generation capability of ChatGPT is robust and raises concerns about student use and academic integrity.3,6-8

 

POTENTIAL BENEFITS FOR FACULTY AND STUDENTS

Artificial intelligence-based tools may be used to help simplify routine faculty administrative, course preparation, and assessment tasks. ChatGPT content generation could be used to create syllabi and class exercises, generate multiple choice quiz questions, or make recommendations for specialized education plans. This may provide more time for faculty to focus on issues requiring teaching expertise and personal engagement with students.3,8 ChatGPT can provide a type of self-tutoring for students through its immediate responses to queries and comprehensive summaries of research studies. ChatGPT can also be used to produce low-level multiple-choice practice tests with answers. Students should be made aware that the reliability of the output must be verified, and all citations validated.2,3

 

RECOMMENDATIONS FOR FACULTY

Increasing personal awareness of ChatGPT's strengths and limitations will allow nursing faculty to take advantage of its benefits while controlling its negative aspects.5 Faculty should inform students that ChatGPT query results may be false and/or unverifiable. Students should have clear guidelines for the use of ChatGPT along with information about academic integrity policies and the associated consequences of plagiarism and academic misconduct violations.2,3

 

Strategies to discourage ChatGPT use include assignments that focus on the analysis and creation levels of Bloom's taxonomy, real-time simulations for in-the-moment critical thinking, problem-solving group projects, individual innovation projects using course concepts, or employing a single case study scenario for the whole class.2,7 Additional strategies include reflective practice-based essays to critically analyze clinical experiences, oral presentations, or live classroom discussions.7 Faculty should be aware that ChatGPT-4 can now process and interpret images and video content embedded in quizzes or examinations. Distance examination proctoring may not detect student access to ChatGPT while test-taking. Artificial intelligence detection tools have shown variable success, so it is best to follow institutional recommendations for software use.2,3,7

 

IMPLICATIONS FOR NURSING EDUCATION

Student temptation to rely on ChatGPT content for learning may decrease their active engagement with course materials and other valid content sources. This may prevent development of skills necessary to personally appraise, analyze, and synthesize data. Nursing students need the ability to critically think and generate independent and creative thoughts to address difficult clinical problems.2,6,7 ChatGPT was designed to mimic human intelligence; however, it may not be able to "solve nursing problems that demand wisdom, critical thinking, intuition, and personal judgment."6 Ongoing research is needed to determine the impact of ChatGPT on nursing student learning outcomes, faculty teaching practices, and ethical standards.7

 

IMPLICATIONS FOR PATIENTS AND PUBLIC HEALTH

Consideration should be given to the potential impact of ChatGPT on patients as healthcare consumers. Patients may query ChatGPT to answer health-related questions much as they used the internet during the COVID-19 pandemic. Nursing students should have a working knowledge of ChatGPT including its limitations in providing accurate public health information for their patients.9 Unfortunately, ChatGPT may write plausible-sounding but incorrect answers to health queries by the public. It is postulated that large-language model-generated misinformation could be rapidly disseminated via social media to create an "AI-driven infodemic" public health phenomenon.10 It is important for nurses to remain informed about AI and help lead policy efforts to ensure the risks and benefits of this evolving technology serve the interests of patients and the public.

 

ChatGPT is an innovative and widely used example of the power of AI. The nature of AI technology ensures that it will continue to evolve rapidly. Nursing faculty will need to remain familiar with the capabilities of new versions as they emerge. There should be open communication between faculty and students about the potential use and misuse of ChatGPT and ongoing dialog about its impact on teaching, learning, and patient health.

 

References

 

1. National Artificial Intelligence Act of 2020. 2020. Cited 2023. https://www.state.gov/artificial-intelligence/. Accessed May 19, 2023. [Context Link]

 

2. Sun GH, Hoelscher SH. The ChatGPT storm and what faculty can do. Nurse Educ. 2023;48(3):119-124. [Context Link]

 

3. Lo CK. What is the impact of ChatGPT on education? A rapid review of the literature. Educ Sci. 2023;13(4):1-15. [Context Link]

 

4. OpenAI. ChatGPT (March 14 version) [Large language model]. 2023. https://chat.openai.com/chat. [Context Link]

 

5. Sallam M. The utility of ChatGPT as an example of large language models in healthcare education, research, and practice: systematic review on the future perspectives and potential limitations. medRxiv 2023.02.19.23286155; doi: https://doi.org/10.1101/2023.02.19.23286155. [Context Link]

 

6. Abdulai AF, Hung L. Will ChatGPT undermine ethical values in nursing education, research, and practice? Nurs Inq. 2023;e12556. [Context Link]

 

7. Choi EPH, Lee JJ, Ho MH, Kwok JYY, Lok KYW. Chatting or cheating? The impacts of ChatGPT and other artificial intelligence language models on nurse education. Nurse Educ Today. 2023;125:105796. [Context Link]

 

8. De Gagne JC. The state of artificial intelligence in nursing education: past, present, and future directions. Int J Environ Res Public Health. 2023;20(6). [Context Link]

 

9. Archibald MM, Clark AM. ChatGTP: what is it and how can nursing and health science education use it? J Adv Nurs. 2023. [Context Link]

 

10. De Angelis L, Baglivo F, Arzilli G, et al. ChatGPT and the rise of large language models: the new AI-driven infodemic threat in public health. SSRN. 2023;11:1166120. [Context Link]