In the world of artificial intelligence, a token is the smallest unit of information processed by an AI model. When a user asks ChatGPT, DeepSeek, Gemini, or Doubao a question, the system does not read sentences like a human. It breaks words down to tiny pieces called tokens. These are the building blocks that allow AI to calculate, predict, and generate responses.
A country, an organization, or an individual’s adoption of AI can be measured by the number of tokens consumed. In March, China’s average daily token consumption exceeded 140 trillion, a thousandfold increase in just two years.
Just how many tokens are needed to produce an AI response? For example, a single one-minute video generated by ByteDance’s Seedance 2.0 model consumes over 4 million tokens at 1080p resolution. A single query can potentially burn hundreds of thousands of tokens. Multiply that by hundreds of millions of daily users, together with the advance of AI technology, and the 140-trillion figure was just a checkpoint along China’s journey with exponential growth expected in the coming months.
Recently, China’s regulators formally designated the Chinese name for a token as ciyuan, combining the character for “word” with the character for “element”. The naming is deliberate. Just as the industrial age standardized the kilowatt-hour for electricity and the container twenty-foot equivalent unit for global shipping, the AI age now has its own standard unit of account.
The National Data Administration identified ciyuan as the “anchor of value” and the “settlement unit” connecting technological supply with commercial demand. With a publicly recognized standard of measurement, an ecosystem of token-based industries is starting to emerge.
Hong Kong occupies a unique position within this enormous token economy. The current national plan explicitly positions the Northern Metropolis as a premier ecosystem for technological research and development, driving “AI+” industrial transformation and cross-boundary laboratory collaboration. Hong Kong’s role must be to serve as a superconnector between Chinese mainland productive capacity and global capital markets in three ways
In the United States, the token economy has developed along a corporate-led path. Jensen Huang of Nvidia describes tokens as a new commodity, comparable to oil or electricity in previous industrial revolutions. OpenAI, Anthropic, and Google all price their application programming interface calls per token. The strategic goal for American AI firms is to build massive data centers that churn out these digital units at high speed to power everything from medical research to automated customer service.
What’s in it for the Hong Kong Special Administrative Region?
Under the 15th Five-Year Plan (2026-30) and the national “AI+” initiatives, the central government has made AI adoption a formal priority across manufacturing, logistics, healthcare, and finance. Token usage is now concentrated in sectors with high information density and fast product cycles: The tech industry leads, followed by consumer electronics, finance, new retail, and business services.
Hong Kong occupies a unique position within this enormous token economy. The current national plan explicitly positions the Northern Metropolis as a premier ecosystem for technological research and development, driving “AI+” industrial transformation and cross-boundary laboratory collaboration. Hong Kong’s role must be to serve as a superconnector between Chinese mainland productive capacity and global capital markets in three ways.
First, Hong Kong gives global investors confidence. The city’s common law system, stable regulatory environment, and international legal frameworks offer a reliable platform for global investors to fund AI projects tethered to the mainland market.
Second, professional expertise is abundant at scale in Hong Kong. As the mainland produces tokens at an accelerating rate, Hong Kong’s legal, financial, and intellectual property services are essential for managing and trading these assets across borders. No other city combines common law, English language capability, and direct physical integration with mainland infrastructure.
Third, the Northern Metropolis serves as the natural testing ground for mainland AI models in international markets. Many global enterprises remain cautious about deploying mainland-hosted models directly. Hong Kong’s data privacy framework, international connectivity, and the dedicated R&D ecosystem make it the ideal bridgehead where mainland large models can be tested, benchmarked, and integrated into multinational workflows.
Hong Kong’s common law system and international legal standards are genuine advantages, but questions of token ownership, cross-border data liability, algorithmic accountability, and dispute resolution for token-based transactions remain largely unanswered. Without updated legislation and clear judicial guidance, investors may hesitate to commit capital, and international partners may have legal and compliance concerns to integrate with Hong Kong-based token platforms. The legal infrastructure must keep pace.
High-volume AI tasks consume tokens rapidly, and the data centers behind them, the token factories, require substantial land and enormous amounts of electricity. Chief Executive John Lee Ka-chiu has consistently promoted the development of Hong Kong’s data center infrastructure in the policy addresses, most notably at the Sandy Ridge site. Building the city’s own capacity is the right long-term decision, but it will take years to bear fruit. Hong Kong needs access to affordable and reliable token-factory capacity now, and cross-boundary arrangements with Guangdong’s data-center corridor are the logical interim solution. Realistic energy pricing and clean power sourcing on both sides of the boundary will determine whether that integration actually works.
The opportunity is real. So are the constraints. The HKSAR government can consider giving the newly established Hung Shui Kiu Industry Park Co Ltd a mandate to develop strategies, perhaps ringfencing a modified version of the top talent program or the Investment as Entrepreneurs pathway, to attract investors and founders of industries capable of helping Hong Kong take advantage of the token economy. Priority sectors should include AI model training, high-density data center operations, token-based settlement, and cross-boundary AI service integration.
Tokens may be small, but their contribution to Hong Kong’s economy is enormous.
The author is a senior lecturer in the Department of Marketing, the Hang Seng University of Hong Kong.
The views do not necessarily reflect those of China Daily.
