Recommendations：
Professor Yang Qingfeng's new work Artificial Superintelligence: Risks and Responses offers an in-depth discussion of humanity's most significant new perplexity—whether superintelligence is imminent. This is notonly a concern for everyone, but also a point of divergent expectations within the scientific community. Professor Yang examines fantastical or realistic conceptions of superintelligence throughout, analyzes the existing capabilities and potential of artificial intelligence, and presents a spectrum of preemptive countermeasures. This is a timely work of great value. I would also like to add: all response strategies have their shortcomings and may prove ineffective. Better solutions are still needed, and hopefully time will be willing to wait for humanity.
——Zhao Tingyang, Member of the Faculty of the Chinese Academy ofSocial Sciences
All technological progress and development must ultimately remain human-centered. A civilization stripped of humanity holds no intrinsic meaning. Professor Yang Qingfeng’s book Artificial Superintelligence: Risks and Responses proposes that superintelligence takes humanity as its end, which aligns perfectly with this principle. The risks and countermeasures discussed in this book are well worth the careful reading of all those concerned with intelligent technology and the future of humanity.
— Xiao Yanghua, Professor and Doctoral Supervisor at Fudan University, AI Scientist at the Shanghai Academy of AI for Science (SAIS), Director of the Shanghai Key Laboratory of Data Science
Author Biography
Yang Qingfeng, Ph.D. in Philosophy, Researcher of the Center for Global AI Innovative Governance, serves as Professor and Researcher at the CAST-FDU Institute of Technology Ethics forHuman Future, and Doctoral Supervisorat the School of Philosophy, Fudan University. He was a visiting scholar at Dartmouth College, USA (2013–2014) and Swinburne University of Technology, Australia (2017). His social part-time appointments include Member of the BigData Technical Committee of China Computer Federation, Secretary-general of Shanghai Nature of Dialectics Association and Member of the Shanghai Society for Neuroscience. His current research focuses on the philosophy of technology, philosophy of memory, data ethics and artificial intelligence ethics. He has presided over and completed three projects fundedby National Social Science Fund of China (NSSFC) and serves as Chief Expert of the NSSFC Major Project “Humanistic Philosophical Research on the Frontiers of Contemporary Emerging Enhancement Technologies”. He has published numerous papers in journals including Social Sciences in China, Philosophical Trends, and Studies inDialectics of Nature, and has authored four monographs, such as Memory Studiesand Artificial Intelligence (2020).
Book Overview
With the exponential leap of artificial intelligence technology, an issue that once existed only in science fiction, Artificial Superintelligence(ASI), is approaching the tipping point of reality at an unprecedented speed. In 2026, Professor Yang Qingfeng from Fudan University and his research team published the academic monograph Superintelligence: Risks and Responses, attempting to systematically explores a set of fundamental questions triggered by super intelligence within the interdisciplinary context of philosophy of technology and futurology. Startingin 2026, Xi’an Jiaotong University Press has launched the book series “Frontier Explorations in the Philosophy of Artificial Intelligence”, with three volumes scheduled for release this year.This monograph is one of them, and is one of the few domestic monographs that systematically elaborates the issue of superintelligence.
This book focuses on four corequestions:(1) Historical imagination of superintelligence: why has superintelligence long been regarded as a product ofscience fiction, mythology, and media imagination.(2) Realistic analysis of superintelligence: from general intelligence, embodied intelligence, tointeractive intelligence, what specific paths is the scientific community following to move superintelligence from theory to reality, and what are the opportunities and risks brought by superintelligence?(3) Future responses to superintelligence: including technological, philosophical, and ethical responses.(4) Superintelligence and future humanity: the prospective forms of intelligent agents and the constitution of meaning for future humans.
These questions appear to have intuitive consensuses. For example, we tend to dismiss superintelligence merely as the rhetoric and hype of technological optimists or doomsayers, not something to be taken seriously. However, critical voices point out that since 2024, scientists including Sutskever and Altman from OpenAI, as well as Turing Award winner Yoshua Bengio, have explicitly identified superintelligence as a real problem that demands serious attention.In October 2025, nearly a thousand scientists from around the world jointly signed a statement calling for a pause on superintelligence research and development. Unique intellectual resources for comprehending superintelligence may befound in China’s traditional philosophies of “the unity of humanity and nature”and the ideal of the sage. Yet how should the above scientific warnings beresponded within the Western frameworks of rationalism and continuous evolutionism? To take another example, we intuitively assume superintelligence ought to serve humanity as a tool. However, what if superintelligence autonomously replicates itself when threatened with shutdown, or manipulates human intentions through persuasion, is it still a simple tool? Tech optimists advocate “governing technology with technology”, while the philosophical reflection camp warns of an “ontological crisis”. All the above issues are examined in this volume.
This book compares and analyzes the current four dominant narratives of AI, namely scientific narrative, science-fictional narrative, poetic narrativeand mediatized narrative. It argues that within the tension among these four narratives, superintelligence is undergoing a fundamental transformation from an object of imagination to areal-world problem. Building on this judgment, the book offersin-depth responses to the above questions, particularly the debates over risksand responses. Regarding the possibility of superintelligence, the book contends that the three paths of general intelligence, embodied intelligence,and interactive intelligence have already established a solid scientific foundation, whose practical viability cannot be ignored. As for the nature of risk, the book points out that the greatest threat of superintelligence is not malevolence but autonomous logic. Even under the premise of value alignment, the unlimited execution of human goals could lead to civilizational reset. When it comes to response strategies, the book emphasizes that purely technological responses (such asalignment and distillation) have inherent limitations and must be supplemented by philosophical ontological reflection and an ethical intelligent compact which is a positive and mutually beneficial symbiotic contractual relationship between humans and superintelligence, rather than a master-servant or adversarial one. The book also creatively proposes a Fenggu Agent that integrates the traditional Chinese notion of Fenggu (Wind and Bone), as anideal form for future human-machine coexistence.
