Operationalizing and Implementing Pretrained, Large Artificial Intelligence Linguistic Models in the US Health Care System: Outlook of Generative Pretrained Transformer 3 (GPT-3) as a Service Model

• Published in: JMIR medical informatics Vol. 10; no. 2; p. e32875
• Main Authors: Sezgin, Emre, Sirrianni, Joseph, Linwood, Simon L
• Format: Journal Article
• Language: English
• Published: Canada JMIR Publications 10.02.2022
• Subjects: Application programming interface; Artificial intelligence; Bias; Cloud computing; COVID-19; Deep learning; Design specifications; Electronic health records; Hospital systems; Language; Life expectancy; Machine learning; Medical screening; Moons; Natural language; Operating costs; Solar system; Text messaging; Viewpoint; Web portals; United States > US; chatbot; generative pretrained transformer; clinical informatics; natural language processing; artificial intelligence

Generative pretrained transformer models have been popular recently due to their enhanced capabilities and performance. In contrast to many existing artificial intelligence models, generative pretrained transformer models can perform with very limited training data. Generative pretrained transformer...