For most of history, if you wanted something written down, you paid a professional. A contract, a letter, your own will: all of it passed through the hands of a scribe. Reading and writing was a trade, and a market. Then, over a few centuries, everyone learned to read. The market for scribes did not become more competitive, and it did not become more efficient. It vanished. Nobody would call that a loss. This essay is about a strange idea. The idea that the scribes’ story is not a curiosity but a template. That true empowerment does not come from better markets. It comes from dissolving them.
Most often, our default mode of thinking about technological innovation is in terms of efficiency and utility. Better optical fibers enable faster data transmission; more energy-efficient chips enable cheaper AI inference. What this framing overlooks are the second-order effects, the things an innovation does to the distribution of power. History offers plenty of examples. Cheap, mass-produced firearms distributed power: A warhorse and a suit of armor were things only a lord could afford. But a rifle was something a farmer could own. The telegraph did the opposite. The network was costly to build and therefore owned by few, enabling the first truly centralized nation-states and continent-spanning corporations. In each case, the direction of the shift was determined less by the technology itself than by the terms of access to it: who could afford it, who owned the infrastructure, who held the knowledge of how it worked.
Which is to say: how a technology is shared with the world is just as important as the technology itself. An innovation can be kept private, locked behind trade secrets, proprietary infrastructure, and exclusive access, or it can be shared openly. That choice determines which asymmetries it creates and which it dissolves, and whether it reinforces existing hierarchies or disrupts them. This, in my opinion, is the ultimate measure of whether a technology truly serves as a democratizing force or merely strengthens established players. But to see why, we first need to examine what markets fundamentally are, which asymmetries sustain them, and which ones corrupt them: secrets, property, capital, and, increasingly, control of the networks that move information itself.
A market is a self-organizing system where participants exchange goods, services, or resources. Markets persist as long as asymmetries exist between participants. Without asymmetries, for better or worse, markets naturally dissolve. These asymmetries come in fundamentally different kinds: some reflect what a participant can contribute, others what a participant can exclude others from. The division of labor is the archetype of the first kind. By specializing, some participants can provide goods or services that others lack the skills to produce. From a computer science perspective, this specialization resembles knowledge sharding: instead of every individual storing the same set of useful ideas (or “memes”), knowledge is distributed across a subset of people, enough to sustain and reproduce the collective knowledge of the community. In this sense, markets act as emergent systems that optimize the storage and use of knowledge within a community.
While some asymmetries, like division of labor, can benefit the entire community, others are less cooperative. For instance, a participant might deliberately create an information asymmetry by keeping knowledge of a new technology secret. By leveraging this trade secret, they could offer a product or service at a lower cost or higher quality than competitors, accumulating wealth in the process.
Consider a company with exclusive access to an advanced AI architecture that can reason more effectively than publicly available systems. This company could offer services with superior contextual understanding that competitors simply cannot match. The exclusive control of this cognitive architecture creates a significant market advantage, which allows the secret holder to extract premium fees while other market participants struggle to compete with inferior technology. Over time, this accumulated wealth can be used not only to control the labor of others but, as we will see, even to influence their attention and decision-making processes. In this way, the closed-source philosophy concentrates power in the hands of the secret holder while systematically disadvantaging other market participants.
Free-market maximalists might argue that competition alone sufficiently addresses this issue, but this perspective overlooks a crucial reality: while competition can temporarily reduce prices, competition without deliberately dismantling exploitative asymmetries inevitably leads to monopolization. It is worth remembering why markets outperform central planning in the first place. As Friedrich Hayek argued in his 1945 essay “The Use of Knowledge in Society”, no central authority can ever assemble the knowledge dispersed across millions of minds. Markets win because prices communicate that knowledge, letting participants act on information they do not themselves possess. Hayek’s mechanism does not ask anyone to surrender what they know. Participants always know different things; that is the division of labor, and it is what gives them something to trade. Prices do their work as long as many independent participants act on their own fragments of knowledge, and as long as every fragment remains contestable by anyone willing to learn. What a trade secret attacks is that plurality. The problem is not uneven knowledge, which markets require. The problem is an asymmetry engineered never to be contested, one that starves the price signal of the competing actions that give it meaning.
Return to the company with the superior AI architecture. Its advantage does not merely let it outcompete others on price or quality; it lets it progressively prune the market itself. Competitors who cannot match the technology lose customers, then revenue, then the ability to fund the research that might close the gap. This reveals something fundamental: markets dissolve at both extremes of asymmetry. With no asymmetries, as noted earlier, there is nothing to exchange, since every participant could produce whatever they need. With an asymmetry maxed out, a single participant becomes the only possible provider, and what remains is no longer exchange but dictate. A market only exists in the band between perfect symmetry and total asymmetry, and unchecked competition around an entrenchable advantage drives it out of that band. The closed-source philosophy is therefore fundamentally in conflict with genuine empowerment, as it increases and entrenches asymmetries rather than reducing them.
Secrets are not the only exclusionary asymmetry. In societies with greater complexity, property asymmetries emerge, where participants own land, real estate, intellectual property, or copyrighted material. As long as an authority, like a government, exists to enforce property rights, property owners can effectively impose a tax on others for access or use. When property holders extract value without contributing labor, this behavior is known as rent-seeking. Many property asymmetries essentially rely on exploiting other participants, weakening and undermining their autonomy. Patent trolling exemplifies this exploitative form of property rights: these entities acquire patents not to produce or utilize the inventions they protect, but solely to extract settlements through litigation. By wielding patents as weapons, they divert resources from genuine research and development toward legal battles, creating no value while extracting wealth from productive companies. Similarly, when individuals or corporations own land or resources they neither developed nor maintain, yet collect payments merely for access, they are engaging in rent extraction that contributes nothing to collective prosperity. This does not mean, however, that all property asymmetries are inherently exploitative. Property rights that stem from genuine labor, such as inventing something new or creating original content, serve as incentives for innovation, and their holders may even share their work openly while expecting compensation for their effort. The key distinction lies in whether a property right exists to protect the fruits of labor or merely as a tool to extract value from others without contributing productive effort.
Whether they originate from secrets or from property, these asymmetries share a common product: accumulated capital. And capital is the most fungible asymmetry of all, one that can be converted into nearly any other. An entity with sufficient capital can sell at a loss until its competitors are eliminated, then raise prices at will, a tactic known as predatory pricing. It can buy up patents, land, or the startups that might one day threaten it. But capital also excludes passively, as table stakes. Some technologies carry fixed costs so high that only incumbents can participate at all. A semiconductor fab is not kept exclusive by any secret; building one is simply beyond nearly everyone’s reach. This is the one asymmetry that openness alone cannot dissolve: publishing the blueprints of a fab democratizes nothing, because openness distributes the knowledge but not the means. It marks the limit of what distribution mechanisms can achieve by themselves. Countering it requires intervening further upstream, at the level of design itself: consciously minimizing a product’s dependency graph, in materials and supply chains alike, and favoring locally available inputs even when that means modifying the design rather than adding another dependency.
But most consequentially, capital can purchase asymmetries of control: ownership of the platforms through which information itself flows. Social media platforms like X (Twitter) or TikTok can influence the formation of opinions by selectively amplifying certain content through their recommendation algorithms, effectively determining which voices reach millions and which remain unheard. Marketplaces like Amazon and search engines like Google can use their proprietary algorithms to influence product and search result visibility, creating winners and losers based on criteria that remain largely hidden from users and sellers alike. The asymmetry no longer operates on what participants can buy or build, but on what they see, believe, and want.
This dynamic, where power comes not from owning things but from controlling the networks that move information, has a name: vectoralism, a term coined by McKenzie Wark in A Hacker Manifesto; Yanis Varoufakis more recently described the same dynamic as cloud capital in Technofeudalism: What Killed Capitalism. Unlike previous examples where market asymmetries emerged from what entities held (secrets, property, capital), vectoralists create asymmetries by owning, operating, and controlling information networks and platforms themselves. By reprogramming participants’ perceptions and decisions through algorithmic control of information flows, vectoralists maximize their control while minimizing others’ autonomy. While traditional asymmetries often center around controlling supply, vectoralism allows platform owners to shape demand itself by determining what users see, when they see it, and how it’s presented, effectively manufacturing desire rather than merely fulfilling it. This distinction has crucial implications for how such power can be disrupted: while a secret or IP can be rendered obsolete by developing alternative technologies, vectoralist power requires the more challenging task of shifting established networks to decentralized alternatives. Breaking the cycle of centrally controlled networks demands novel algorithms with specific properties that enable the transition from platforms to protocols, a shift Mike Masnick has argued for in his essay “Protocols, Not Platforms”.
All of this points back to the idea we began with, a realization that runs against the grain of most thinking about technology and markets. The goal is not to create more markets (as the current enthusiasm for prediction markets and the financialization of everything would have it), nor even, necessarily, to make existing markets more efficient. Recall that markets persist as long as asymmetries exist between participants. When knowledge is widely distributed, a good cheaply produced, or a service easily replicable, markets dissolve. Remember the scribes: universal literacy dissolved their market, and nobody calls that impoverishment. True economic empowerment means intentionally reducing non-labor-based asymmetries so that the productive capability of every participant rises: weakening trade secrets, ending rent extraction, breaking network control, and, where capital itself is the barrier, designing for local reproducibility. Each asymmetry dissolved this way converts a fee into a skill, a dependency into labor productivity. This is what the open approach ultimately aims at: not better markets, but more capable people.