The Infinite Alphabet of Knowledge: How Knowledge Grows, Diffuses, and Gains Value

Overview

This talk reframes the history of science as the history of knowledge itself, showing how knowledge grows, moves across space, and gains value in economies and societies. It weaves together theories of learning curves with real world stories about big experimental projects, from the mythical Yachai to the diffusion of early industrial technologies.

Key threads include how learning is driven by practice, the diffusion of knowledge through people and networks, the non fungible and highly specific nature of knowhow, and how policy can harness these principles through targeted investments and management of endowments.

Takeaway

Readers will come away with a framework for assessing knowledge growth, diffusion, and value, and with concrete examples of how to design policy and institutions that align with these principles.

Introduction: Reframing Knowledge as an Alphabet

The talk begins by reframing the history of science as a history of how knowledge transforms from colloquial words into formal concepts. It sets up a triptych of principles that govern the growth, diffusion, and valuation of knowledge, arguing that these principles are essential for understanding why ambitious knowledge projects sometimes fail when they ignore them. A central narrative thread uses the Yachai project in Ecuador as a cautionary tale: a billion-dollar effort to create a city of knowledge in a former plantation near Quito, intended to host biotech, robotics, AI, and a supercomputer, all integrated with a university and a modern city. The enthusiasm at launch was high but eventual outcomes showed many of the classic mistakes of equating knowledge creation with architecture and infrastructure alone. The speaker uses this story to motivate a systematic study of how knowledge actually grows, moves, and earns value across scales and contexts.

Time and Learning: The Growth of Knowledge

The first major section dives into the quantitative study of knowledge growth over time. It begins with Leon Louis Thurston, who in the early 20th century constructed a learning curve from a 1916 Pittsburgh typing class, showing that performance improves rapidly with practice, then saturates. This curve is characterized by the idea that learning is a function of experience rather than time. Theodore Paul Wright later rediscovered a similar effect in airplane manufacturing costs, and Leonard Rapping extended this to the wartime construction of Liberty ships, reinforcing the idea of a power-law or similar learning curve across activities and tasks. A crucial twist emerges with Gordon Moore’s transistors, where the learning is exponential and tied to price reductions and rapid production capacity expansions. The narrative emphasizes that these curves are not universal constants; they reflect different generations and technologies that build on each other. Transistors evolved from mesa to planar designs and ultimately to integrated circuits; the diffusion of transistor technology was accelerated by a network of people, capital, and institutions, and not simply by a single team or a linear improvement path. The story is rounded with an analogy of lighting in the UK: from candles to gas to electricity, with each generation offering new light at lower costs, illustrating a relay of generations that pushes the boundary of what is possible. The core takeaway is that knowledge grows by a series of fast early gains driven by practice, which then slow down, but at the same time enabling new generations to leapfrog previous limits. Even so, forgetting remains a force. The talk notes empirical evidence that knowledge decays if it is not actively practiced and renewed, summarizing this as a regime where learning must offset forgetting in order for knowledge to persist and evolve.

Diffusion: The Geography of Knowledge Movement

The second major segment centers on diffusion and the geographic movement of knowledge. It begins with the Samuel Slater narrative, emphasizing that diffusion is not merely about the transmission of a single technology but about the movement of tacit knowhow that depends on people, networks, and the ability to implement a technology in a new locale. Slater’s migration to the US and his collaboration with Moses Brown sparked the first automated cotton spinning in America, illustrating that knowledge is not fungible or easily transplantable; it requires skilled labor, context, and the institutional support that allows tacit expertise to take root. The story expands to Masaru Ibuka and the concept of absorbent capacity—some places have greater ability to absorb and utilize external knowledge due to existing research ecosystems and technical ecosystems. Ibuka’s postwar work on magnetic tape exemplifies how a country or region can reuse and repurpose knowledge by combining literature review, empirical tinkering, and local supply chain creativity to reproduce a technology. The lecture underscores a key point: knowledge diffusion is geographically constrained but highly mediated by social networks and the mobility of people who carry tacit knowledge. A provocative aside addresses non fungibility: the idea that knowledge is highly specific, and moving from one domain to another may require distinct skill sets and materials. The Vespa case illustrates this: an aircraft engineer, faced with postwar constraints, pivoted into scooter design by reusing core engineering knowledge in a new, closely related domain. The Irreducible point is that diffusion is shaped by the geography of knowledge: not just where is knowledge created but where it can be absorbed and applied given related activities in proximity. The discussion then broadens to the concept of the product space, where networks connect products, technologies, and skills that tend to co-occur in export baskets, patents, and skill sets. The relatedness principle is introduced: the probability of entering a new activity grows with the number of related activities present, while leaving an activity becomes less likely with more related activity presence.

The Infinite Alphabet: Specificity and Non Fungibility of Knowledge

A central conceptual move is the idea of knowledge as an infinite alphabet. Knowledge is incredibly specific and non fungible; one piece of knowledge is not simply added to another to produce a sum. A telling anecdote about a Florida lawyer named Charlie illustrates how extremely specialized knowledge can become a highly valuable service. The tale of Charlie, who extends cases for clients for a fee and builds a niche business around a tight, context-specific capability, demonstrates that a small, highly specialized piece of knowhow can be extremely economically valuable in a given environment. The talk then returns to the broader conceptual framework: diffusion happens along networks of related activities, and knowledge may leap across domains only when there are meaningful related activities to anchor the transfer. The Vespa and related cross-impacts among aircraft, motorcycle, and scooter design serve as a powerful example of non fungible, highly specific skills that can be ported to a closely connected domain, but not to unrelated areas without heavy retooling and redesign. These stories emphasize that to understand diffusion we must embrace the infinite alphabet concept and recognize the essential role of relatedness in enabling diffusion across activity spaces.

Product Space, Relatedness, and the Geography of Knowledge

The talk formalizes the diffusion into a geography of knowledge, a network where nodes are products, technologies, or knowledge domains and links reflect co export, shared workers, or co appearance in patents and publications. Across these networks, a consistent pattern emerges: the probability of entering an activity increases with the number of related activities present and the probability of exiting decreases with relatedness. This is the principle of relatedness: diffusion is not random but constrained by knowledge geography. This reframing allows us to see knowledge diffusion as a process that unfolds in both tangible spaces (cities, labs) and in the intangible spaces of ideas and methods. The infinite alphabet perspective remains central: the diffusion from one related domain to another depends on the compatibility of underlying competencies and tacit knowhow. The conclusion of this section ties diffusion to policy implications: design and policy should cultivate environments where related activities are densely connected so diffusion can occur more readily and with less friction.

Valuing Knowledge: The Third Law and Quantifying Potential

The third law asks how we estimate the value of knowledge by counting the pieces of the infinite alphabet present in a location. The author uses a thought experiment comparing Walt Disney and Pepo, Chile’s Condorito creator, to illustrate how a similar genius with different sets of available alphabets (creative resources, talent pools, partners) ends up with different trajectories. Disney’s move to Los Angeles, where there were abundant resources for voice acting, animation, and storytelling, created a richer alphabet for him to exploit, pushing him toward a series of expansions (animation, feature films, theme parks, real estate development) that defined a new stage of the knowledge economy. Pepo, based in Santiago, faced a much thinner alphabet, limiting the scope of his enterprise. The maps of economic complexity and relatedness are referenced as tools for policy design, illustrating how content and strategy can be tuned to maximize growth potential and diffusion. The talk moves from general theory to practical policy: how to measure a location’s potential by partialing out the effects of specialization, and how to plan investments that create enduring, scalable knowledge ecosystems rather than one-off mega projects.

Case Studies and Policy Lessons: Yachai and Zhongguangzhou

The Ecuadorian Yachai example is revisited to draw practical lessons. The original plan allocated large sums to infrastructure with the hope that the university would naturally generate high-tech industries in the city. The speaker then outlines an alternative approach: instead of building a single, money-intensive project in a remote area, invest in an endowment model that sustains high quality research through chairs rather than ephemeral funding. An endowment payout, for example, at 4%, could support 40 chairs at $1 million annually each in perpetuity, top-tier labs, and computational work, without creating the half-finished infrastructure problem seen in Yachai. The CPU university in Ecuador’s San Francisco de Quito is presented as a counterexample of organic growth: a private university founded through perseverance, collaboration, and slow, steady development rather than a large, centralized push. The narrative shows that the most successful knowledge initiatives in Ecuador have been driven by committed teams and organic growth within the existing ecosystem rather than grand, isolated megaprojects. The Zhongguangzhou example, by contrast, demonstrates how a city-centered, policy-enabled cluster—an avenue of entrepreneurs with guided funds that combine public investment with private incentives—can dramatically raise venture capital activity and stimulate rapid cycles of technological development. This bifurcation illustrates two axioms: (1) architecture alone does not guarantee success; (2) alignment of endowment-based funding, short cycle technologies, and dense local networks can produce faster returns and resilient development outcomes.

Architectural vs Organic Innovation: Architectural Innovation as a Threshold

A crucial theme is architectural innovation, the kind that requires rethinking how components fit together rather than incremental improvements on an existing architecture. The jet engine vs propeller aircraft example, Blockbuster vs Netflix, and the digital transformation of consumer distribution all illustrate that architectural changes redefine the entire system; underestimating such shifts invites disruption and reduces incumbents to laggards. The discussion emphasizes that the most successful players recognize these shifts early and adapt by reorganizing around new architecture rather than attempting to retrofit the old model. The lesson for policy is to spot architectural shifts early and provide support for the new ecosystem rather than propping up obsolete structures.

Infinity and the Alphabet: Final Reflections

The talk ends on a philosophical note: while the universe is finite, the alphabet of knowledge is effectively infinite due to combinatorial possibilities. Even with a finite set of letters in each person’s hand, the collective acts of collaboration create an ever-growing space of possibilities. The closing reflection argues that knowledge’s value lies in its dynamic potential to create new combinations, new technologies, and new ways of organizing society. The speaker presents knowledge as a living, evolving system that requires human agency, collaboration, and careful policy design to realize its potential.