Artificial Intelligence—From Starting Pilots to Scalable Privilege
Ce dossier présente deux études évaluant la performance d'agents conversationnels (chatbots) alimentés par l'intelligence artificielle pour délivrer des informations concernant le traitement de certains cancers (mammaires, pulmonaires, prostatiques, colorectaux et cutanés)
The medical world is now enamored with our newest potential helper, large language models (LLMs) and in particular chatbots, such as ChatGPT, a service offering from OpenAI. While artificial intelligence (AI) is still considered the broadest term, with a computer mimicking aspects of human intelligence, an LLM (such as a chatbot) is a type of AI technology that can understand and generate human-like text. Large language models work by being trained on vast amounts of written information (such as that found on the internet) to learn patterns and relationships between words and sentences. In the medical sphere, generative pretraining transformer (GPT) models are already at or near passing performance on the US Medical Licensing Examination. Generative pretraining transformer models are being evaluated for assisting in the generation of responses to inbox messages from patients in electronic health record systems, while other LLMs can predict 30-day all-cause readmission, in-hospital mortality, length of stay, and even insurance denials.