The Cross-Domain Limitations of Tech Leaders
As the AI boom continues, some technology leaders have put forward bold but debatable ideas: traditional money may weaken, and tokens (Token) will become the “compute currency” of the AI era because they are standardized, tradable, and in high demand, capable of directly measuring AI output and pricing services. Elon Musk and others emphasize that energy is the true “currency” of the AI era — AI data centers are like roaring factories, and whoever controls compute power and electricity controls value creation. There are also views that AI skills will become the new global currency, while data and attention serve as the exchange for “free” services. In an extreme post-scarcity scenario, “work becomes optional, money becomes irrelevant,” and the economy may shift to direct allocation of physical goods or computational resources.
These ideas sound imaginative but reveal the common limitations of cross-domain thinking. Tech experts excel in the depth of AI, but in monetary finance they often fall short. Cryptocurrency and monetary economics require vast interdisciplinary knowledge; a purely technical perspective easily overlooks the fundamental attributes of money.
The Fundamental Rules of Monetary Economics
The underlying logic of money lies in scarcity — money cannot be issued arbitrarily, or it will lose trust. This can be understood through the classic Fisher Equation:
MV = PQ
- M: Average money supply in circulation
- V: Velocity of money (average number of times money turns over per unit time)
- P: General price level
- Q: Total volume of transactions in goods and services (or real output)
This equation is essentially an identity: the left side represents total monetary expenditure (how much money is spent and how fast), while the right side represents the total value of goods and services transacted (what price things are sold at and in what quantity). In the quantity theory of money, if V (velocity) and Q (transaction/output volume) are relatively stable in the short term, an increase in M (money supply) will directly push up P (price level), causing inflation; the opposite may lead to deflation. Central banks can influence velocity by adjusting interest rates, thereby keeping money supply roughly aligned with the real economy. This is the core of the quantity theory.
Any “new money” must pass this test — whether it can maintain supply-demand balance and value stability.
Why Compute Power and Tokens Are Not Money
Tokens (Token) are indeed a key metering unit in the AI era (every model call consumes tokens), but they are essentially unlimited “production inputs,” similar to oil or electricity in the industrial age, rather than money. They lack the intrinsic scarcity of gold and have no credible issuing institution like fiat currency. Model providers can expand supply indefinitely according to demand. How can you ensure the issuance precisely matches economic growth? And how can it serve as a reliable “yardstick” rather than a volatile commodity?
Money typically needs to fulfill three attributes simultaneously:
- Store of Value: Long-term preservation of value. Gold, Bitcoin, and dollar reserves possess this; tokens depreciate with model iterations and compute expansion, making them unsuitable.
- Medium of Exchange: Convenient for payments. Gold and Bitcoin are weaker than the dollar due to volatility; tokens are mainly used for internal settlement and lack broad acceptance.
- Unit of Account: A stable pricing standard. The dollar is strongest here; gold and Bitcoin are weaker due to volatility; tokens function more like variable costs than a fixed standard.
The idea of direct allocation of computational resources is a regression to physical economy thinking from thousands of years ago. In complex modern economies, it has long been replaced by monetary systems — physical allocation struggles with coordination, incentives, and scalability.
The Global Currency in the AI Era: Stablecoins Remain the Best Hope
The most realistic and promising solution remains dollar-anchored stablecoins (especially USDC and USDT). Their value is directly pegged to the dollar. The Federal Reserve, as an independent institution, issues in a relatively transparent manner, and the dollar holds the highest credit among fiat currencies. Despite the dollar’s flaws, it remains the best anchor in the current global environment. Stablecoins can be seen as the optimal bridge from traditional finance to the AI economy.
At the 2026 World Economic Forum in Davos, Circle CEO Jeremy Allaire stated clearly that “tens of billions of AI agents” will emerge in the next three to five years to conduct continuous economic activities, and they will need an efficient payment system. “At present, there appears to be no viable alternative other than stablecoins.” Traditional fiat payments rely on bank intervention, clearing, and manual review, often taking days — far too slow for the high-frequency micro-payments required by AI agents, such as autonomously purchasing compute, subscribing to APIs, or paying for data.
Blockchain-based stablecoins achieve instant settlement: the ledger balances automatically through consensus mechanisms, with strong programmability, making them especially suitable for machine-to-machine transactions. Coinbase’s Payments MCP function allows models like Claude and Gemini to access wallets, complete fiat on-ramps, and execute stablecoin payments directly via natural language instructions, without complex API keys. This significantly lowers the barrier for autonomous economic activity by AI agents.
Future Directions: Institutional Optimization of Dollar Stablecoins and Decentralized Evolution
Although dollar-backed stablecoins have clear advantages, they also have institutional shortcomings (such as reserve transparency, dispersed issuers, lack of unified coordination, and no revenue-sharing mechanisms). Hong Kong has favorable regulatory and financial conditions and is poised to capture a share of the dollar-anchored stablecoin market. However, the licensing process for Hong Kong dollar stablecoins has been delayed beyond the original schedule. Whether to consider issuing USD-denominated stablecoins directly in the future remains an issue worth watching. For more details, refer to the “Market Commentary” section on the Chainless website (chainless.hk).
In the long term, decentralized stablecoins (such as over-collateralized DAI and projects like DW20) are expected to become more resilient supplementary tools in the AI era. They operate entirely on blockchain, execute automatically via smart contracts, and do not rely on a single issuer, better supporting decentralized finance (DeFi) and autonomous coordination among AI agents. However, at present, decentralized stablecoins still lag behind fiat-anchored ones in scale, stability, and regulatory acceptance. The future will require finding a balance between compliance and innovation.
Conclusion
The AI era will not completely overturn money but will accelerate its digitization and programmability. Tokens, compute power, and energy remain important infrastructure and commodities, but they cannot replace money that possesses scarcity, stability, and broad acceptance. Stablecoins — especially dollar stablecoins deeply integrated with the AI agent economy — are most likely to become the bridge connecting traditional finance and the intelligent economy.
Tech leaders’ imagination deserves respect, but monetary issues must ultimately return to the basic laws of economics. Who holds a leading position in stablecoins will gain the upper hand in the global monetary system of the AI era. This evolution has only just begun and is worth continuous observation. Decentralized stablecoins are likely to become one of the important directions in the future.
