Embracing Artificial Intelligence

Abstract

We are at a threshold moment for shaping the nature of medical care in the near and distant future. A plethora of ethical frameworks, guidelines, and principles exist for artificial intelligence (AI), inspiring us to learn more.1 A recent lecture encouraged me to experiment with AI tools, increase my AI literacy, and eventually integrate AI into my work and personal life.2 AI is here to stay, and our challenge is to understand and utilize this technology to the best of our capabilities. Wendy Moncur, a cybersecurity researcher at the University of Strathclyde in Glasgow, United Kingdom, reminds us: “If you’ve got a tool [such as AI] that’s going to make things more efficient, then it makes sense to use the tool. But we need information to help us use the tools responsibly, and in a way that allows us to do good science.”3 As journal editors, we help authors navigate uncertainty, communicate data that can inform practice, and call for more research to fill gaps in the literature and in practice. To do this, we must continue to read extensively, but also utilize all modes of modern technology. As editors, we guide authors as they turn data into knowledge and disseminate work that responsibly guides practice to improve patient outcomes. We embrace even those things that are unclear to us, such as AI. Nurse leaders, researchers, professional organizations, specialty societies, and individual professionals are also working hard to move science forward using this new technology, which is accelerating faster than most of us can comprehend. We know that AI offers tremendous potential to solve many issues burdening the workforce. Yet, much is unknown, and fears and skepticism abound. As with all new innovations and technology advances, it is important to follow expert advice and search for valid evidence. Generative AI (such as ChatGPT) and big-data methods are helping us gain new insights. As I leap into AI, a first step has been untangling the acronyms and understanding basic meanings.3 I have learned that AI is an umbrella term to describe how machines or computers mimic human logic. Large language models (LLMs) are a subset of AI. These models are specific to generative AI, which means they are created by machine learning, which becomes more sophisticated each day describing inputs as moving closer to deep learning using neural network architecture. LLMs such as GPT-4 arise from Open AI and are trained on massive data sets. These models do not have human capabilities in the traditional sense. They cannot read, sense emotions, show emotions, or provide critical thinking. Instead, they are trained on text data broken down into tokens (words or parts of words). LLMs generate text based on patterns learned from their training data, predicting the next token in a sequence. As I learn this new language and definitions, I have come to understand the applications that will change our workflow. This month, I will participate in a training program offered by my organization, who recently announced the adoption of Copilot, a Generative AI tool that provides a data-protected environment within the academic community. In this workshop, we will learn to incorporate Copilot into teaching. I hope to learn about its unique features compared with other AI tools such as ChatGPT and Gemini and receive guidance on promoting responsible use among students and authors. The workshop promises to provide hands-on experience with Copilot AI, exploring its potential for student use and assignment evaluation. This is an exciting opportunity as Copilot for Microsoft 365 add-on was made available only to all Microsoft 365 business customers on January 15, 2024 (https://copilot.microsoft.com). Innovative leaders anticipate major improvement in productivity using products such as Copilot with Teams. Plans to use Copilot with Excel, PowerPoint, Outlook, and SharePoint will be part of our workflow. I hope to learn more each day. As nurse leaders, we must be involved in defining the optimal balance between human effort and AI assistance. AI-powered solutions are being offered for every application. This technology has already shown us that the status quo of human effort alone is not sustainable. These are just a few of the things I have learned about this technology. Al shows promise to perform ambient documentation such as summary notes from flowsheets; provide prediction analytics that may uncover droves of data that will help predict falls, pressure injury, and catheter-associated urinary tract infection; synthesize literature; draft patient education materials; titrate medication via algorithms; create troubleshooting guides for infusion pumps and other medical devices; create policies for things like massive blood transfusion; AI-powered solutions may also help us figure out staffing, supply chain issues, and other workflow-related issues; unleash unlimited potential for wearable technologies; guide mobile devices for patient care and communication. Dr Sachin Kheterpal reminds us that many health care conditions are silent, such as hypertension, but wearable technologies and AI-assisted data management may provide early indicators to some clinically significant data that may be buried in massive data files. He reminds us that AI is uniquely poised to do things that are beyond our human limitations, time, and often limited of practice scope. Certainly, within the complexities of modern health care and the onslaught of information coming to us each day, we need this help. With such promise comes the responsibility to use AI in an ethical manner for the protection of patients. Sims and Cassel introduce many ethical frameworks that may guide the broad societal implications of AI. One recommendation includes applying the core Belmont principles as a conceptual framework for AI in health care.1 Amid my fears and optimism, I will embrace AI in this unique time in health care’s history, I challenge others to join me. Open a free ChatGPT account and experiment with it. Search for reputable information about AI. Enroll in webinars, course, and lectures presented by experts. Engage in discussion about the implications for health care and test AI-powered tools for writing and research.4,5

Ähnliche Arbeiten