Summary of the Key Points
This article focuses on the "Pax Silica" alliance promoted by the United States, highlighting that the global AI competition has evolved from a battle of models and chips between companies to a full-scale confrontation of ecosystems across the entire industry chain. In this context, Hong Kong's opportunity does not lie in taking sides but in leveraging its "one country, two systems" framework to become a "super connector" that links China's AI industry (manufacturing, use cases) with global computing power, capital, and markets. This requires strategic investments in international computing hubs, new types of AI exchanges, and digital industrialization, while also being vigilant against competitors like Singapore. Hong Kong must develop an overall strategy to become a leading international financial center for the AI era.
I. The US's "Pax Silica": Not Just a Chip Alliance, but a New Order for the Entire AI Industry Chain
Many people think "Pax Silica" is just another semiconductor alliance, but it's much more than that. The AI industry chain is far more complex than chips: training large models requires GPUs, which in turn need advanced manufacturing processes; these processes depend on lithography machines and materials, chip production requires energy, data centers need power and cooling, and the commercialization of models depends on cloud computing, data, and capital—forming a lengthy chain from "key minerals → energy → chip equipment → computing power → large models → applications → capital."
The US is bringing together 24 economies, including Japan (materials), the Netherlands (lithography machines), South Korea (storage chips), Australia (minerals), and the Middle East (energy/capital), to create a complete AI economic system. Just as in the oil era, those who controlled the entire chain of oil fields, transportation, refining, and dollar settlements dominated the energy landscape; in the AI era, those who can organize computing power, capital, and technology will have the upper hand.
II. Sino-US AI Competition: From a "Model War" to an "Ecosystem War"
In the past, the competition was about who had more model parameters or a smarter ChatGPT. Now, it's about who has a more complete ecosystem.
- US Advantages: It has top global tech companies like OpenAI and NVIDIA, as well as a developed venture capital and dollar-based financial system that can provide funding and technology.
- Chinese Advantages: China has the most comprehensive manufacturing base in the world, a large pool of engineers, and extensive use cases (e.g., AI in factories, logistics, healthcare). Chinese models follow a "low-cost + open-source" approach, which is more suitable for cost reduction and efficiency improvement in traditional industries.
This competition is not about who wins or loses, but about the superiority of different models: one driven by technology and capital, the other by manufacturing and applications. Hong Kong is positioned perfectly at the intersection of these two systems.
III. Hong Kong's Core Opportunity: to Become an "Irreplaceable Super Connector"
People often ask which side Hong Kong should align with, but the real value lies in its role as a connector.
In the past, Hong Kong made money by connecting Chinese companies with international capital (e.g., mainland companies listing on the stock exchange). In the AI era, it must evolve into a hub that connects China's AI industry with global computing power, capital, and markets:
- Connecting to the mainland: China has the most AI use cases (e.g., in factories and logistics) and needs computing power and capital.
- Connecting to the world: The world has computing resources (e.g., data centers in Singapore), capital (e.g., Middle Eastern sovereign funds), and technology (e.g., Dutch lithography machines).
"One country, two systems" gives Hong Kong a unique advantage: it can deeply integrate with the mainland economy while operating according to international rules—something that Shanghai and Shenzhen cannot match, and something Singapore cannot take away.
IV. Three Key Actions for Hong Kong
1. Build an International Computing Hub: The computing power of Hong Kong's Sha Ling data center will be 36 times what it is now, but it should not just be a repository for servers. Instead, it should transform computing power into tradable assets—e.g., by standardizing the rights to use computing resources for the next decade and making them available for global investment and trading. Hong Kong's financial expertise (in financing and securitization) can play a crucial role here.
2. Develop New Types of AI Exchanges: Traditional exchanges trade stocks, but in the AI era, they will need to trade new assets such as GPU usage rights, data center capacity, and API revenues from AI models. Hong Kong can use tokenization (not just for cryptocurrency) to facilitate these transactions, turning real assets into digital forms.
3. Redefine Industrialization: While traditional factories are costly in terms of land and labor, Hong Kong can focus on "digital industrialization"—using computing power as infrastructure, data as a production factor, and models as tools. For example, the value created by an AI data center could exceed that of a traditional industrial park, and the capital flow generated by an AI trading platform could be greater than that of dozens of factories.
V. Hong Kong's Competitors and Strategy
Hong Kong's main competitors are not Shanghai or Shenzhen, but Singapore, Dubai, and Seoul, all of which aim to become key nodes in the global AI network. Singapore is investing in data centers and the Southeast Asian market; Dubai is leveraging its energy and capital for AI development; Seoul relies on Samsung's storage chips.
Hong Kong must avoid adopting a traditional financial center approach (e.g., just building data centers or providing loans to tech companies). Instead, it needs a comprehensive strategy that integrates resources such as Sha Ling computing power, the stock exchange, tokenized transactions, and大湾区 use cases to create a closed loop of "capital → computing power → models → industry → assets → capital."
The window for action is narrow; waiting for the global AI order to settle will be too late. The ultimate question for Hong Kong is: in the AI era, should it remain a traditional financial center or a super hub that connects the global AI industry?
Discussion Topic: Which direction do you think Hong Kong should prioritize? A computing hub, tokenized asset trading, or an AI industry capital allocation center? Feel free to leave your comments and share your thoughts!