The article discusses the limitations of Large Language Model (LLM) powered chatbots and introduces a new system, ExploreLLM. This system aims to reduce the cognitive load on users by structuring complex tasks, exploring different options, and allowing for more personalized responses. The paper presents a user study that shows the benefits of using ExploreLLM for planning and exploratory tasks, suggesting a future where users interact with LLMs beyond the traditional text-based chatbot format.
Publication date: 1 Dec 2023
Project Page: https://arxiv.org/abs/2312.00763v1
Paper: https://arxiv.org/pdf/2312.00763
