Trending Applications of Large Language Models: A User Perspective Survey

作者：Yiqun Zhang, Mingjie Zhao, Yunfan Zhang, Yiu‐ming Cheung · 发表于：IEEE Transactions on Artificial Intelligence · 年份：2025 · DOI：10.1109/tai.2025.3620272 · 被引用次数：4 · 研究领域：Topic Modeling

Large Language Models (LLMs) have revolutionized industries, learning, and life, becoming an indispensable tool in daily digital interactions and reshaping how humans access knowledge and creativity. Despite the ubiquity of LLMs, a critical gap persists between non-Artificial Intelligence (AI) background users and overly specialized literature. Most existing LLM surveys mainly provide technical principles, engineering details, and professional evaluation, rather than conveying intuitive and easy-to-follow content to serve ordinary users. This paper, therefore, is positioned as a user-centric and jargon-free review for all learners and practitioners who are curious about LLMs or intend to use them. Uniquely structured around organizational (e.g., text summarization) and innovative (e.g., content generation) application scenarios, and the user roles of engineers, designers, and supervisors, this survey demystifies LLM capabilities through intuitive taxonomies, case studies, and hands-on evaluations. Specifically, eighteen mainstream LLMs are statistically selected as representatives to answer the following questions that ordinary users may be interested in: 1) Where do LLMs come from, what are they, and why are they so smart (Section I); 2) Which application scenarios are suitable for certain LLMs and why (Section II); 3) How users with different roles use LLMs (Section III); and 4) What do specific applications of LLMs look like (Section IV). In addition, readers will gain act...