https://ianreppel.org/from-steam-to-silicon/ Ian Reppel home search insights fiction about 30 May 2025 * Ian Reppel * 4 min From Steam to Silicon: Patterns of Technological Revolutions From agriculture to AI, each major economic revolution is driven by a new kind of value conversion: a way of transforming one type of value into another. These revolutions are not defined by a single invention, but by a sequence of innovations: a core conversion, the infrastructure to scale and distribute various innovations, and the technologies that reduce the cost of distance and delays. We shall analyse six major revolutions through this lens and show how each laid the groundwork for the next. Technological revolutions There are seven key components to each technological revolution: 1. Core conversion: a new capability that transforms one form of value into another. Value can be either tangible (e.g. natural resources and goods) or intangible (e.g. trust, information, and actions). 2. Scaling infrastructure: physical systems that make the conversion repeatable and cost-efficient at scale. 3. Spatiotemporal compression: tools that shrink the cost of distance or time in the economy by i) increasing connectivity, ii) reducing the latency of interactions, or iii) expanding the scale or speed of economic coordination. 4. Key resources: physical or non-physical ingredients that are needed in the core conversion recipe. 5. Economic mode: each technological revolution has its own dominant way of creating and capturing value. 6. Centralization/decentralization arc: core conversions start centralised to exploit economies of scale, but as complementary technologies lower transaction costs, production disperses. 7. Administrative innovations: standards and bureaucratic tools to support the technological revolution at scale and across geographical areas, which may or may not be from the era itself. Each technological revolution is also a bridge to the next revolution, in that its innovations enable the next one to thrive. Timeline In the table below, I have summarized the pattern for the agricultural, financial, and four industrial revolutions: The ((core conversion)) is made efficient thanks to ((scaling infrastructure)) to turn ((key resources)) into ((economic mode)) products and services, first centrally in ((centralization arc)) and later decentralized in ((decentralization arc)). The economy is made more interconnected and reduces distance or time between interactions because of ((spatiotemporal compression)). This is all made possible because of supportive ((administrative innovations)). More details on the four industrial revolutions (IRs) can be found in Late to the Revolution. Core Scaling Spatiotemporal Economic Centralization/ Administrative Revolution conversion infrastructure compression Key resources mode decentralization innovations Enables next arc * plant/animal * settlements (temple) * writing Agricultural labour - domestication * trade routes * crops surplus granaries - * accounting Surplus enables (ca. 10 000 food * irrigation * roads * livestock (food) feudal land * standardized cities and BCE) * plough tenure measures specialization * printed money state-chartered * bookkeeping Capital markets credible * bills of exchange promises (of banks and stock * codified fund massive Financial (ca. trust - institutions (e.g. (transferable future capital exchanges - bond commercial law infrastructure 1600) credit banks) IOUs) payments) markets and * pooled risk initiatives * clearing systems merchant credit (IR1) networks Country-wide industrial logistics IR1 heat - * railroads * coal products factories - patents (with needed for (1760-1840) motion steam engine * canals * iron (e.g. workshops / time-limited distributed textiles) factory hubs monopolies) power production (IR2) Electrification * assembly lines * oil mass * standardized time and global IR2 chemical - * electrification * telecommunication * steel consumer power plants - (time zones) logistics (1870-1914) motion * ICE * road networks * copper goods (e.g. cars (ICE) * limited-liability enable mass appliances, companies media and cars) automation (IR3) * internet (i.e. * silicon * digital Networked satellites, * rare services computing IR3 electricity telecommunication) earths * mass digital protocols enables (1950-2000) - electronics * globalization * aluminium consumer mainframes - PCs (e.g. TCP/IP, APIs) intelligent information (i.e. airplanes, * energy goods autonomous container ships) systems (IR4) Automated * mobile internet * data * algorithmic/app decision-making IR4 information AI (4G+) * compute autonomous cloud - edge marketplaces enables (2010-present) - action * (I)IoT * energy agents * serialization planetary-scale formats (self-) optimization Prior to the agricultural revolution, (decentralized) family farms were the norm. It is common for a technological revolution to centralize the core conversion initially to make it more efficient and scalable. Afterwards, the conversion is dispersed, such as with settlements connected to the same temple/palace granaries through trade networks and much later tenancy of land in feudal economies. Why it matters Each core conversion is a textbook general-purpose technology (GPT). Most mainstream models describe what happens, but not why; they focus on the diffusion of specific technologies, but not the functional shift those technologies cause. This model helps us ask: * What kind of conversion do we bet on next? * What infrastructure must be built to scale it? * What barriers in space and time need to be collapsed? * What are the primary resources and who controls these? * How will the economic mode shift once it becomes mainstream? * Where are we in the centralization/decentralization arc and how can we disperse its benefits better? * What administrative tools are required to support it? The answers to these questions help us not only explain the past, but anticipate what kind of futures are possible, and what each requires to succeed. The model can also guide policy decisions: instead of wondering what the next gadget is, we can invest wisely in scaling infrastructure, tools for spatiotemporal compression to ensure the benefits from the core conversion technology are spread far and wide, and appropriate regulation to ensure interoperability. The pattern of initial centralization and later decentralization is clear across all six technological revolutions, which means we can anticipate that after the initial infrastructure investments of today, we may need planetary sensor grids to ensure we can tackle challenges such as climate change appropriately. Modern medicine, antibiotics, and vaccines have added many years to global life expectancy. While that is an immense welfare gain, it does not constitute a new technological revolution in this framework: there is no new value conversion. Each medical breakthrough still relies on the prevailing infrastructure and logistics. Medical gains do amplify revolutions, as healthier people work longer, learn more, and innovate faster. Where does quantum fit? Quantum computing sits inside IR4 as a supportive technology accelerator, similar to GPUs. For specific computations in cryptography, chemistry, machine learning, and optimization, quantum computers are expected to deliver massive speed-ups. Certain intractable problems may only be solvable on quantum devices in the future. This may lead to novel materials and pharmaceuticals, provide significant performance boosts in training AIs, or improve global logistics and supply chains. Like steam in the 1770s and semiconductors in the 1950s, quantum computing is still in the lab, but it will be transformative once it is ready for the real world, which is about 5-10 years from now. What's next? Every revolution is won or lost on the boring stuff: standards, infrastructure, and fair access. Steam needed rails, electricity needed wires, and IR5 will need a planetary-scale layer of data and compute that is open, resilient, and reachable by everyone. Policy makers have three levers that have already been proven: 1. Mandate open, royalty-free standards for data, APIs, and edge hardware: interoperability cuts compliance costs and boosts competition. 2. Treat broadband and smart grids as 21st-century railroads: a 10 percent increase in broadband penetration bumps GDP by up to 1.4% and smart grids have a 10x ROI. 3. Build digital public infrastructure to flatten inequality: instant payments slash merchant fees and pull millions into the formal economy. So, invest in standards, digital infrastructure, and parity now, or spend the next decades paying a coordination tax that no technology can offset. [?] Buy Me a Coffee Related posts Late to the Revolution: The High Cost of Tech Lag 29 May 2025 The Penalty of Being Human 01 May 2025 Leadership Lessons from Jazz, Improv, and Japan 12 Mar 2025 Late to the Revolution: The High Cost of Tech Lag Back to top