Integrating Expertise in LLMs: Crafting a Customized Nutrition Assistant - CHI 2024

Explore a conference talk that delves into the development of a customized nutrition assistant using Large Language Models (LLMs). Learn how researchers collaborated with registered dietitians to evaluate GPT-4's capabilities in providing accurate and personalized nutrition information. Discover the mixed-methods approach used to validate LLM outputs and the resulting design guidelines for prompting nutrition-related information. Gain insights into the creation and refinement of "The Food Product Nutrition Assistant" prototype, tailored for generating detailed food product descriptions and personalized nutrition advice. Understand how dietitian-reviewed template instructions enhance the quality and accuracy of LLM-generated nutrition information, potentially revolutionizing the way people make informed food choices.