Complex systems science in the AI era: a pivotal paradigm for scientific research

Abstract Reductionism has underpinned modern science since the 17th-century Scientific Revolution. This methodology decomposes systems into minimal units and deduces wholes from parts, succeeding across physical, life and social sciences. However, reductionism reveals limits as the scientific frontier shifts from identifying building blocks to understanding how components generate collective behavior. We possess massive data and precise local equations but fail to predict cell fates, financial crises or cognitive emergence in neural networks. Complex systems science—a mid-20th-century discipline exploring the structures, behaviors, evolution and laws of complex systems—integrates with diverse fields to transform scientific paradigms. Rapid AI development triggers this paradigm shift. AI algorithms handle massive data and complex systems, while AI systems themselves constitute complex systems whose progress depends on complex systems science. In this NSR Forum, we convene five researchers to discuss the essence of complex systems science and its integration with other disciplines to support a pivotal scientific paradigm shift in the AI era. Tingting Gao Postdoc Researcher, Network Science Institute, Northeastern University, USA Wei Lin Professor, School of Mathematical Sciences and Research Institute of Intelligent Complex Systems, Fudan University, China Yu Liu Associate Professor, Department of Systems Science, Beijing Normal University at Zhuhai, China Jiang Zhang Professor, School of Systems Science, Beijing Normal University, China Lei Guo (Chair) Professor, Academy of Mathematics and Systems Science, Chinese Academy of Sciences, China