🕰️ In 1988, when the world was just beginning to embrace personal computers and the internet was the domain of a narrow circle of scientists, two researchers—Mark S. Miller from Xerox PARC and K. Eric Drexler from MIT—published a paper that was three decades ahead of its time. In the collection The Ecology of Computation, they laid out a concept that sounded like science fiction: the computers of the future should operate like a market economy, where software "agents" trade computational resources, pay each other for services, and compete for the "consumer’s" attention.
💡 Their "Berkeley Protocol" (an unofficial name coined later, based on Miller’s workplace) wasn’t code or a standard. It was a manifesto calling for a rethinking of computational system architecture. Instead of centralized planning, where the OS dictates who gets how many resources, they proposed a model of spontaneous order governed by prices and competition. Their agoras (from the Greek agora—market) would become digital entrepreneurs in an environment where CPU time and memory were commodities, and code was capital.
⚖️ The foundation of this digital economy was to be the principle of encapsulation of information, access, and resources. Put simply, every software object would receive inviolable "property rights" over its data, the right to control who could interact with it, and—key innovation—over the computational power allocated to it. This created the basis for trade: an object could "sell" a portion of its CPU time or memory to another object if it was profitable.
🧩 Miller and Drexler drew a direct parallel with the work of Nobel laureate Friedrich Hayek on the role of prices as a mechanism for transmitting decentralized knowledge. "The price of a resource shows how much it’s demanded by the entire system," they wrote. A high price for CPU time signaled to an agent that its task might not be so important at the moment, prompting it to yield resources to a more "valuable" competitor. Thus, local decisions based on the simple rule "sell high, buy low" were supposed to lead to globally efficient resource allocation without any central planner.
💸 For calculations, a currency was needed. The authors envisioned internal banks that would ensure secure transfers of virtual money between agents, including across different machines. They even anticipated multiple currencies and exchange rates between them—a strikingly accurate foreshadowing of modern DEXs (decentralized exchanges).
🤔 So why wasn’t this elegant theory implemented in the '90s or 2000s? The problem wasn’t technology—it was economics, specifically transaction costs. Miller and Drexler themselves pointed to it, citing economist Ronald Coase and his question "Why do firms exist?" In the human economy, market transactions between the smallest actors (like individual employees within a company) would be unbearably expensive due to the costs of negotiation, accounting, and oversight.
⚙️ The same would happen inside a computer. The CPU time spent calculating every kopeck for every byte of memory across thousands of microscopic objects would devour all the benefits of market-based allocation. The authors’ solution was as brilliant as it was practical: scalable application of market mechanisms. At the level of small objects, fast, low-level algorithms (like auctions for task schedulers) would operate, while large, complex software modules—true "digital firms"—could trade with each other just like companies in the real world.
🔗 There’s no direct line from Miller and Drexler’s work to Bitcoin or Ethereum. But their ideas turned out to be eerily prescient. The concept of smart contracts—self-executing agreements with predefined conditions—is a direct implementation of their idea of creating rules that are "genuine constraints" rather than just threats of punishment for violations.
🛡️ An even more important legacy was the theory of object-capabilities (ocap) developed by Miller and his colleagues, which became the cornerstone of security in decentralized systems. It provides the very "access encapsulation" needed for safe interaction between mutually distrustful parties—exactly what blockchain requires.
🌱 Their 1988 work isn’t just a historical curiosity. It’s a fundamental exploration of how decentralized systems can self-organize using economic incentives. Today’s token economy is a special case of their far more general and profound idea: any complex system, whether a society or a computer network, can achieve efficient order not through top-down commands but through voluntary cooperation among independent agents guided by local gain.
🧠 The irony is that the technology that finally made their dream of a global computational agora possible—blockchain—didn’t emerge from university labs but on the fringes of the internet, among crypto-anarchists and libertarians. It makes you wonder: perhaps the true power of breakthrough ideas lies not in their immediate adoption but in their ability to wait until the world is ready to understand them.