Zaidi Rohma, Asghedom Jonathan, Singh Som
University of Missouri Kansas City School of Medicine, Kansas City, Missouri.
University of Missouri Kansas City School of Medicine, Kansas City, Missouri and University of Texas Health Science Center, Houston, Texas.
Mo Med. 2024 Nov-Dec;121(6):495-498.
The growing use of generative artificial intelligence (AI) in the public sphere allows for a greater degree of disseminating information worldwide. For patients, there is a growing body of literature exploring how the generative artificial intelligence models can be used in improving the health literacy of patients, especially in cases of acute pulmonary embolism, where patients require deep, concise understanding of there disease and management. This study measured the readability of the generative responses created by publicly available AI models, and found that ChatGPT, Google Gemini, and Microsoft CoPilot do not currently meet the United States readability recommendations. Given the growing use of these models, future investigation on the longitudinal readability measures may help profile how these generative AI models adapt in their deep learning processes.
生成式人工智能在公共领域的使用日益增加，使得信息在全球范围内的传播程度更高。对于患者而言，越来越多的文献在探索生成式人工智能模型如何用于提高患者的健康素养，尤其是在急性肺栓塞的情况下，患者需要深入、简洁地了解自己的疾病及治疗方法。本研究测量了公开可用的人工智能模型生成的回复的可读性，发现ChatGPT、谷歌Gemini和微软Copilot目前未达到美国的可读性建议。鉴于这些模型的使用日益增加，未来对纵向可读性指标的研究可能有助于了解这些生成式人工智能模型在深度学习过程中的适应性。
