By the scenic Dishui Lake, a profound dialogue on the future of science unfolded as scheduled. On October 23, 2025, as a prelude to the 2025 World Laureates Forum (WLF), the Academic Symposium on Trustworthy Software Intelligence Methods was successfully held at the School of Software Engineering (Dishui Lake International Software Institute), East China Normal University (ECNU). Jointly organized by the Shanghai Lingang Science and Technology Innovation Development Foundation and ECNU, the symposium, under the theme “Science in Future: Shanghai and the World”, focused on the intersection of artificial intelligence and software engineering, bringing together world-leading minds to explore how to build trustworthiness in the age of intelligent science.
This was not only an academic symposium, but also a deep interdisciplinary dialogue bridging disciplines and connecting today’s innovators with tomorrow’s scientific leaders.
A Visionary Leader for an Interdisciplinary Dialogue
The event welcomed a distinguished scientific pioneer — Professor Michael LEVITT, 2013 Nobel Laureate in Chemistry, Vice President of the World Laureates Association (WLA), and Professor at Stanford University.
As one of the earliest scientists to apply computer simulation to chemical research, Professor LEVITT brought unique depth and perspective to the dialogue. From molecular dynamics simulation to intelligent software systems, his groundbreaking multiscale modeling methods not only propelled chemistry into the era of computational research but also offered profound insights into understanding complex systems. Professor LEVITT’s scientific journey itself stands as a vivid testimony to the evolution of modern scientific paradigms.
Also attending were Associate Professor Amir GOHARSHADY from the Department of Computer Science, University of Oxford, and Professor JIANG Xuefeng from the College of Chemistry and Molecular Engineering, ECNU. In an age when artificial intelligence is reshaping research paradigms, these three scholars jointly led a thought-provoking journey across disciplinary frontiers.
Dialogue on “Scientific Paradigms and Responsibility Boundaries in the Era of Trustworthy Intelligence”
The three scientists engaged in an in-depth discussion with attending students around the theme “Scientific Paradigms and Responsibility Boundaries in the Era of Trustworthy Intelligence.” The dialogue was conducted in an open, dynamic atmosphere, sparking vibrant exchanges and insights. Students raised a series of thought-provoking questions from their own research experiences: How can we critically assess the reliability of AI-generated content? In safety-critical domains such as autonomous driving, how should innovation be balanced with system trustworthiness? And how can we construct regulatory frameworks that ensure accountability while respecting privacy? The experts responded from their respective fields: Professor LEVITT emphasized, through examples from chemical experimentation, the indispensable role of empirical validation for AI predictions. Professor JIANG Xuefeng proposed a pragmatic approach in which AI undertakes repetitive work while researchers focus on innovative exploration. Associate Professor GOHARSHADY illustrated the importance of verifiable algorithms in critical systems by drawing parallels between smart contracts and railway control systems.
In the academic presentation session, Amir GOHARSHADY, JIANG Xuefeng, and WANG Peixin, a young researcher from the School of Software Engineering (Dishui Lake International Software Institute), each shared insights from their respective fields, offering multi-dimensional perspectives from theoretical frontiers to practical innovation.
Amir GOHARSHADY presented “Scalable Program Analysis via Decompositions of Control-Flow Graphs.” He analyzed the verification challenges posed by the massive scale and game-theoretic complexity of modern software systems and proposed innovative solutions based on parameterized complexity theory and tree decomposition methods.
JIANG Xuefeng delivered a report titled “AI for Chemistry.” He pointed out that artificial intelligence is profoundly reshaping the research paradigm of chemistry. Using examples of AI-accelerated molecular design and reaction prediction, he illustrated how AI-driven research paradigms enhance reliability, automation, and intelligence in R&D processes across various scientific domains.
WANG Peixin shared his work titled “Structural Abstraction and Refinement for Probabilistic Programs.” His research addresses the challenge of verifying the reliability of probabilistic programs, proposing a novel structural abstraction–refinement approach that provides new methodologies for analyzing software trustworthiness in areas such as differential privacy and randomized algorithms.
As one of the satellite forums of the 2025 World Laureates Forum, this jointly organized academic symposium successfully established a high-level interdisciplinary platform for scientific exchange. Since the founding of the World Laureates Forum in Shanghai seven years ago, Lingang and WLA have resonated in unison. Every autumn, the world’s “brightest minds” gather by Dishui Lake to focus on global scientific frontiers, fostering integration and innovation while contributing wisdom and strength to building an open and trustworthy scientific future.
Source: School of Software Engineering, East China Normal University
