Application of Gemini in Public Health Amid the Artificial Intelligence Era

In December 2023, Google DeepMind unveiled its large language model, Gemini, which is shown to be capable of understanding and processing multimodal information. This article explores the applications of Gemini in public health. Unlike traditional models, Gemini integrates and analyzes information through diverse modalities (text, image, video, etc.) fostering new applications. Notably, Google Gemini is presented as the first model to surpass human experts on the Massive Multitask Language Understanding (MMLU) benchmark, which combines multiple subjects including science, technology, engineering, and mathematics (STEM), humanities, and others to test the knowledge and problem-solving abilities of artificial intelligence (AI) models. While Gemini's capabilities are claimed to surpass established models such as ChatGPT, offering improved accuracy, efficiency, and interdisciplinary research integration, its utilization in research is still limited. A wide range of applications, including healthcare communication and delivery, medical diagnosis and decision-making, early disease diagnosis, public health surveillance, interventions, and education, drug research, and adverse effects prevention can benefit from Gemini's features. By extracting insights from sizable datasets and comprehending complex medical information, Gemini enables researchers and practitioners. However, responsible use with expert oversight is crucial. Overall, Gemini's capabilities offer new opportunities for improving public health research, practice, and ultimately, population health outcomes.