From Gold and Grain to Petabytes and Petaflops: Rewriting the Wealth of Nations

In 1776, Adam Smith published The Wealth of Nations, a key work that defined a nation’s prosperity in terms of production, specialization and tangible assets such as gold, land and crops. For centuries, wealth was quantified by harvested bushels, extracted minerals and manufactured goods. Today, however, the concept of wealth is shifting dramatically. In the era of artificial intelligence (AI), prosperity is more about what a nation can compute, store and learn than about physical possessions. The new measures of power are petabytes of data and petaflops of computing capacity.

This change extends beyond technology; it represents a fundamental shift in civilization. Like the Industrial Revolution, which reshaped economies through mechanization and energy, the AI revolution is transforming them through data and computing. Petabytes, the enormous amounts of data produced by digital interactions, now serve as the essential resource of the modern economy. Meanwhile, petaflops, the immense processing power, drive the conversion of this data into valuable insights, boost productivity and enable new innovations.

At the core of this change is a new understanding of capital. Traditionally, in classical economics, capital meant machinery, infrastructure, and financial assets. Now, data has become capital, and computing power serves as labor. Machine learning models, trained on vast datasets and run on powerful computers, are increasingly taking over tasks once performed by humans, including prediction, classification, optimization and even creativity. In this new landscape, economic value arises not only from manual effort but also from the coordination of algorithms and system architectures.

Yilei Shao, known for her work on Silicon-based Economics, characterizes this change as a civilizational shift in which silicon is replacing the carbon-based industrial system, much as the industrial age replaced the agricultural era. The idea is that intelligence, computation, algorithms, data, models and agents now replace energy as the main factor of production, just as energy once replaced human and animal labor. While the silicon chip is key, the larger shift is deeper.

From the “invisible” hand to the ‘algorithmic’ hand

Adam Smith famously described markets as driven by an “invisible hand,” a decentralized mechanism in which individual self-interest unknowingly advances the common good. In this view, prices emerge from countless human decisions, each shaped by limited information and bounded rationality. Today, that invisible hand is being progressively replaced or supplemented by an algorithmic hand.

In algorithmic markets, humans no longer make all decisions about pricing, allocation, and consumption. Instead, intelligent systems process large datasets in real time to guide these decisions. Dynamic pricing algorithms adjust prices based on demand signals, inventory levels and user activity. Recommendation systems influence consumer preferences by shaping what they see, often before conscious choices are made. Supply chains are optimized not through occasional planning but through ongoing, machine-driven adjustments.

Markets are no longer solely self-organizing; they are becoming engineered systems, crafted and managed by those who develop and oversee algorithms.

This change fundamentally shifts how markets are coordinated. While the invisible hand relied on dispersed knowledge and slow feedback, the algorithmic hand uses integrated data, real-time feedback and predictive analytics. Markets are no longer solely self-organizing; they are becoming engineered systems, crafted and managed by those who develop and oversee algorithms.

The implications are significant. While the AI-driven algorithmic hand can greatly enhance efficiency and reduce information gaps, it also poses risks, including opacity, algorithmic collusion, and the centralization of power. Unlike the spontaneous, unowned invisible hand, the algorithmic hand is intentionally created and managed, prompting critical questions about accountability, transparency and fairness.

From GDP to the silicon index

If wealth changes, our methods for measuring it must adapt. For many years, GDP has been the primary indicator of economic success. However, GDP was designed for an industrial age focused on tangible manufacturing, uniform products and concrete transactions. It struggles to capture the characteristics of a digital economy, where value is often intangible, dispersed and driven by algorithms.

We are thus observing the early stages of a transition from GDP to what could be termed a “Silicon Index.”

The Silicon Index evaluates a country’s economic strength not just by output but by its ability to generate, process, and apply intelligence. It considers factors such as data richness, computing power, algorithmic skills, digital infrastructure and human expertise in AI. Unlike GDP, which reflects past performance, the Silicon Index focuses on a nation’s future potential in an intelligence-driven economy.

This emerging viewpoint aligns closely with Yilei Shao’s idea of “silicon economics,” which views semiconductors, especially silicon chips, not just as production inputs but as the essential foundation of modern economies. According to this perspective, a country’s economic strength is becoming more dependent on its role in the semiconductor supply chain, including design, manufacturing and the integration of semiconductors into AI technologies.

Silicon economics reshapes the classic factors of production.

Shao’s Intelligent Domestic Product (IDP) directly addresses the measurement gap. While GDP tracks goods and services, IDP measures intelligence outputs derived from silicon assets. The two metrics are complementary: the Silicon Index gauges a nation’s intelligence capacity, while IDP reflects actual output, much as industrial accounting separates infrastructure from production. A high Silicon Index with low IDP signals unused capacity; the reverse suggests efficient deployment. Without both, countries risk navigating AI-driven development blind to their position or progress.

Silicon economics reshapes the classic factors of production. In addition to land, labor and capital, compute, connectivity and code play crucial roles. Nations excelling in chip design, fabrication, and high-performance computing exert significant influence over global innovation, supply chains and geopolitical stability. The recent semiconductor shortages highlighted this shift: supply disruptions affected many sectors, including automotive, healthcare, and defense.

In this context, the Silicon Index shifts from a mere metric to a strategic perspective. It emphasizes the need to invest in semiconductor ecosystems, cultivate AI talent, and build resilient digital infrastructure. Additionally, it highlights the dangers of overconcentration, in which a limited number of regions dominate key parts of the silicon economy.

However, a meaningful Silicon Index should extend beyond technical capacity alone. It needs to include considerations of equity, governance and sustainability to ensure that the advantages of silicon economics are widely distributed and in harmony with societal objectives.

Rethinking wealth, power and policy

This development impacts not just measurement but also policy and governance. Countries now compete to establish data centres, secure semiconductor supply chains, and develop national AI capabilities. Just as oil shaped geopolitical influence in the 20th century, data and computing are determining it in the 21st. The emergence of “compute nationalism” indicates an increasing awareness that controlling petaflops is as strategically important as natural resources.

However, this transition may also exacerbate global inequalities. The infrastructure needed to produce and handle petabytes, such as high-speed connectivity, advanced chips and cloud platforms, is not equally available worldwide. The digital divide now extends beyond internet access to access to intelligence itself.

No single country can resolve this, underscoring the need for the UN to develop a new multilateral framework to oversee the global intelligence economy with greater legitimacy than bilateral or regional initiatives.

Shao points out a hidden governance issue: control over the extraction, pricing, and settlement of intelligence. As intelligence becomes essential to production, future disputes will center on establishing terms for its creation and distribution worldwide, rather than traditional resources like oil or rare earths. This situation generates tension between producing and importing nations, reminiscent of past resource dependencies. No single country can resolve this, underscoring the need for the UN to develop a new multilateral framework to oversee the global intelligence economy with greater legitimacy than bilateral or regional initiatives.

To navigate this transition, we should reexamine the principles guiding economic governance. Developing new frameworks for considering data as a shared resource is essential. Investing in computational infrastructure as a public good is also necessary. Additionally, we need to enhance human capacity not just in technical skills but also in ethics, governance and interdisciplinary approaches.

Furthermore, the increase in petabytes and petaflops necessitates a new social contract. As machines take on more cognitive work, the question arises: how do we share the benefits? If algorithms influence economic results, how can we hold them accountable? These are inherently political issues requiring collective solutions.

An environmental aspect also exists. The substantial energy needs of large-scale computation raise sustainability concerns. As digital infrastructure expands, it is essential to align it with climate objectives by adopting efficient hardware, green data centres and optimized algorithms.

Conclusion: governing the new wealth

The shift from gold and grain to petabytes and petaflops is not a break from Adam Smith’s vision but rather its progression. The core idea of The Wealth of Nations was to organize resources efficiently to enhance societal welfare. Although the types of resources have evolved, the fundamental goal still stands.
The concept of the invisible hand is transforming into an algorithmic, AI-powered mechanism. Traditional GDP is now being complemented by new indicators such as the Silicon Index and, as Yilei Shao suggests, the IDP, tools aimed at showcasing and managing the real drivers of prosperity in the 21st century. Economic strength is becoming more rooted in silicon-based technology.

The wealth of nations now lies not in vaults or fields, but in data flows and computational systems. The focus is no longer just on how wealth is generated, but on how it is encoded, processed, and regulated.