C.A. Micchelli, W.L. Miranker
Journal of the ACM
Generative AI have immense potential in digital health, potential to enable scalable, personalized interventions and surface novel insights. Large Language Models (LLMs) like GPT-4 and LLaMA-3 have shown unprecedented performance in understanding, generating, and reasoning with natural language. However, hallucination is a big concern for using LLMs to directly provide medical information. In this talk, we will explore 3 case studies to demonstrate some innovative ways of using LLMs in digital health that have low-risk yet yield high-reward.