A Paradigm Shift in Service Research: The Case of Service Composition

Recent advancements in artificial intelligence, particularly in machine learning and neural networks, have significantly influenced various domains, including service computing. Large Language Models (LLMs) are at the forefront of this transformation, introducing new paradigms for automation and decision-making. This paper examines the evolving impact of LLMs on service composition, a fundamental problem in service computing. By analyzing shifts in research approaches, methodologies, and system architectures, we highlight how LLM-driven automation challenges traditional composition techniques. The discussion provides insights into emerging opportunities, limitations, and research directions, emphasizing the need to rethink service composition in the era of AI-driven automation.