“In the early days of the internet, many ventures adopted money-burning models and went through trial-and-error phases. Speculative bubbles are inevitable. When these bubbles subside, a portion of the capital invested become sunk costs, which will in turn lay the groundwork for subsequent large-scale industrial deployment,” said LIU Dong, Director of the China Future Internet Engineering Center.
As one of the pioneers in China to engage with the internet industry since the 1990s, he shared his perspectives on China’s internet development and the vision for “Next-Generation Internet” during the 2025 Shanghai Forum, while unveiling the inherent connection between internet, artificial intelligence and data flow.
Q: How do you perceive the evolution of China’s internet landscape and the concept of “Next Generation Internet”?
Liu: The internet fundamentally represents interconnectedness, enabling information sharing that transforms work patterns, social governance and diverse aspects of the world. In the 1990s, the embryonic internet emerged in China alongside accelerating globalization. This convergence suggested that future economic and technological integration would require a widely accessible, interoperable network—favoring open standards and open-source models over closed, proprietary systems.
Initially, the internet’s business models were unclear, leading to widespread experimentation with volunteer-based or free frameworks—a phase resembling today’s AI development. These challenges spurred international cooperation to establish shared standards and infrastructures. Later, the mobile internet’s rise, coupled with China’s unique market conditions, dramatically expanded its influence, accelerating industrial upgrading and economic growth. Today, as the internet ecosystem permeates daily life, its benefits—from digital finance to education—have become universal.
By the late 1990s, IPv4’s limited address space and scalability issues proved inadequate, especially with the rapid expansion of the Internet of Things (IoT). IPv6 adoption became imperative for infrastructure upgrades, serving as a cornerstone of the “Next Generation Internet”—a broader vision encompassing enhanced speed, security, and interoperability. IPv6’s vast address space and improved efficiency form the bedrock for future innovations like the metaverse (requiring seamless device connectivity), AI (dependent on distributed data processing), and blockchain (benefiting from enhanced node communication). This transition represents a massive upgrade across industries—from smart cities to decentralized networks—unlocking untapped innovation potential.
Q: China leads in IPv6 deployment but still faces compatibility challenges. What are the key strategies to overcome application-layer adaptation bottlenecks? When can ordinary users expect to experience the tangible benefits of IPv6?
Liu: The internet’s performance and user experience are fundamentally shaped by the capabilities of its transmission media, such as satellite networks, which determine connection speeds. Historically, the internet operated on a “best-effort” basis, but this approach is no longer adequate—today’s demands require near-flawless performance. With enhanced infrastructure, the internet can deliver richer services, driving more users to invest in bandwidth. This growing financial commitment creates the substantial data flow necessary for the industry to thrive.
As IPv6 becomes ubiquitous, application-layer compatibility follows naturally—every networked service, including apps, now requires IPv6 support.
For ordinary users, there’s no lag in adapting to IPv6; they’re seamlessly immersed in this environment, often unaware of its underlying mechanisms. These elements are already in place. Looking ahead, the incremental advancements—such as more intelligent agents—remain uncertain. We cannot yet predict what forms of AI, built on future models, will join the network.
Q: You proposed that China should become the “World’s Data Factory”. What strategic significance does this vision hold amid global value chain restructuring, and how can China systematically advance this goal?
Liu: High-quality data is the cornerstone of future AI and industrial applications. China’s vast engineering talent pool—ranging from elite system architects to grassroots technicians—gives us a competitive edge in high-end industries and foundational data tasks like annotation, processing, and mining. By centralizing high-quality data aggregation, China could spark a new industrial revolution.
Wherever unique applications emerge, high-quality training datasets will follow. However, publicly accessible training data may dwindle as intellectual property (IP) protections tighten—a paradox that could stifle AI advancement unless governance frameworks strike a balance between innovation and IP rights.
Q: As a scientist witnessing China’s transition from tech follower to leader, what core competencies should youth cultivate in this transformative era? What advice can you offer to Fudan students?
Liu:Fudan University’s recent reforms—emphasizing open-mindedness and the fusion of technology with humanities—will be pivotal in addressing AI’s governance challenges.
Today’s youth have grown up as beneficiaries of the internet, but they must now become pioneers of AI, just as earlier generations pushed the boundaries of digital technology. To lead in this era, I urge students to embrace the culture of innovation—experiment fearlessly and think beyond existing frameworks, and to master AI tools proactively—start as skilled users, then deepen their understanding to develop visionary insights about the future.
(END)
Writer: FENG Zihan
Proofreader: WANG Jingyang
Editor: WANG Mengqi, LI Yijie
