On April 26, Professor Sun Ninghui, an esteemed expert in computer system architecture and an academician at the Chinese Academy of Engineering, presented a crucial lecture titled “The Development of Artificial Intelligence and Intelligent Computing” to the Standing Committee of the National People’s Congress. Despite receiving limited coverage from Chinese media, the full text of the speech, published online, provided a rare and insightful glimpse into Beijing’s AI aspirations.
Born in 1968 in Shanghai, Sun Ninghui has played a pivotal role in China’s technological landscape. His impressive career includes leading the National Intelligent Computer Research and Development Center and serving as the dean of the School of Computer Science and Technology at the University of Chinese Academy of Sciences.
Sun’s lecture went beyond academia; it served as a strategic roadmap for China’s AI goals and concerns. He underscored AI’s dual nature, acknowledging its potential for significant technological progress while also highlighting security risks, especially concerning politically sensitive information. AI-generated deepfakes, fake news from tools like ChatGPT, and the spread of misinformation through fake news sites underscored AI’s potential to erode social trust. For Beijing, these are not just technical issues but strategic vulnerabilities that the leadership is eager to address.
Sun’s emphasis on AI’s capacity to produce politically sensitive or “incorrect” information was notable. He pointed out that the AI landscape is filled with inaccuracies, biases, and susceptibility to manipulation – characteristics that could be exploited by adversaries. This is a significant concern for China, driven by the government’s focus on controlling politically charged content. Beijing has implemented strict regulations to ensure AI development aligns with approved standards.
The tech rivalry with the United States was another key theme in Sun’s lecture. With superior talent, foundational algorithms, and computational power, the U.S. leads in AI. China’s current status as a fast follower is one that Beijing is eager to change.
Sun argued that U.S.-imposed restrictions on high-performance computing products and advanced semiconductor technologies pose significant challenges. Bans on advanced chips like the A100 and H100 hinder China’s AI progress, showcasing the strategic control the United States maintains over China’s tech ambitions. This impact particularly affects companies like Huawei and SMIC. Sun suggested that China should innovate domestically and broaden supply chains to reduce dependence on U.S. technology, aligning with Beijing’s goal of securing its high-tech future.
Sun’s assessment of China’s underdeveloped domestic AI ecosystem was revealing. Compared to NVIDIA’s robust CUDA ecosystem, China’s AI tools and talent pool are still nascent, highlighting the need for a comprehensive approach to create a competitive AI environment. The lack of synergy across AI technology layers poses a significant challenge that Beijing aims to overcome through substantial investments and strategic initiatives.
He outlined three strategic pathways for China’s AI development. The first involves aligning with U.S.-led systems, a practical but limiting approach given current geopolitical tensions. The second is constructing a closed, proprietary system suitable for specific sectors but lacking scalability and global reach. The third, which Sun deemed most promising, involves embracing an open-source model through global collaborations like RISC-V. By championing a collaborative global ecosystem, China aims to break existing monopolies and facilitate entry for domestic enterprises, aligning with Beijing’s vision of becoming a global tech leader.
Investing in new infrastructure is a cornerstone of China’s proposed AI strategy. Sun stressed the importance of robust data and computational infrastructure, including establishing national data hubs, developing foundational AI models, and integrating computational resources nationwide. The goal is to leverage data as a national asset, making AI services as accessible and essential as utilities like water and electricity. This strategy underscores Beijing’s belief in the necessity of a strong infrastructure to sustain growth, whether traditional or AI-driven.
A significant difference between China and the U.S. in AI strategy is China’s emphasis on the real economy, focusing on AI’s practical applications in industries like manufacturing. While the U.S. drives innovation in virtual sectors, China aims to enhance its manufacturing sector through AI integration. Deploying AI in traditional industries while fostering innovation in emerging sectors is crucial for China’s economic growth and global competitiveness, according to Sun.
Despite Sun’s optimism and broad vision, Beijing’s AI ambitions face challenges. The government’s control over politically sensitive content may hinder innovation and restrict open scientific discourse. Relying on state-approved values for AI development raises ethical concerns and the potential for misuse. While China’s strategy to overcome U.S. restrictions through domestic innovation shows promise, the ability to achieve necessary technological breakthroughs in the face of global competition remains uncertain.
Sun’s lecture was not just an academic exercise; it was a strategic declaration as Beijing aims to shape a comprehensive and forward-looking AI strategy. This rare insight into the leadership’s mindset offers a critical perspective on how China plans to leverage AI to maintain its competitive edge in the tech arena.