Every few decades, a new technological architecture arrives that changes the way economic activity itself can be organized. The internet did this to information. Today, convergence of software-defined resources, ubiquitous sensing and connectivity, and increasingly capable AI is creating conditions for a simultaneous rearchitecture of economic activity across almost every major industry.
Recent technology cycles have produced a world in which market power is concentrated in a small number of distribution platforms. In the coming great rearchitecture, we need to take important lessons from the last 30 years.
Three Reasons the Internet Went Wrong
The Internet offered a similar opportunity. A very different outcome from what was imagined was influenced by a few key choices that compounded over time to produce the economy we have today, where economic value is highly concentrated in a small number of distribution platforms:
1. The Architecture Encouraged Silos.
The internet went wrong in part because its architecture, which for security reasons constrained the ability of sites to interact with each other and users to control what happens on sites, encouraged the creation of isolated silos of content rather than an open, interconnected ecosystem in which users and creators could retain meaningful control. Each site became a destination, a walled garden with its own rules, identity, data, and economic model, but sites proliferated so rapidly that most had relatively little power in a sea of options. Instead of empowering individuals to own their digital identities, relationships, and creations, the dominant architecture concentrated control in the intermediaries that hosted, organized, and monetized them. Users became dependent on platforms for access to their social networks, audiences, and accumulated content, while creators increasingly found themselves building businesses on rented infrastructure, subject to changing algorithms, fees, and terms of service. The internet connected the world, but its architecture often left the people who created its value with little power over how that value was used.
2. Free Software Negated IP.
The free software movement contributed to this outcome through an important but ultimately incomplete conception of freedom. By emphasizing freedom to use, modify and share software, it challenged restrictive intellectual property regimes, in many cases expressly negating rights that creators would otherwise have to control their creations. This made software more accessible and collaborative, but the widespread embrace of software as something that should be free also created an environment in which the economic value of intellectual property was frequently dismissed or treated as an obstacle to progress. Large platforms could incorporate open-source software, build enormous businesses on top of it, and capture the resulting economic value without compensating the creators of the underlying technology. The movement’s ideals were ultimately co-opted by companies whose business models appropriated the work of others at scale. In this sense, fear of legal monopolies, which led to the rejection of proprietary rights over software, did not produce freedom; instead, it produced natural monopolies, shifting market power downstream to platforms that controlled distribution, infrastructure, data, and access.
3. Free Market Ideology.
The third failure was in market structure and public policy. As predicted, the internet unlocked incredible network effects. In a period during which both political parties favored free markets, lax antitrust regulation allowed dominant platforms to acquire competitors, control essential infrastructure, and leverage their advantages across adjacent markets. Concentration grew unchecked. The result was an internet in which a small number of companies exercised extraordinary influence over search, social interaction, commerce, advertising, cloud computing, mobile ecosystems, and the flow of information. That concentration has extended beyond the internet itself, as the largest technology companies have become central players in the broader economy and increasingly important gatekeepers to opportunity.
What began as a technology that promised to distribute power has evolved into an economic system in which power is concentrated among a few intermediaries. The combination of siloed architectures, weakened intellectual property rights, and permissive market structures has left creators and users with too little control, while allowing the companies that control the platforms to capture a disproportionate share of the value created by everyone else.
The Great Re-Architecture
The current technological revolution may ultimately be remembered less for the individual AI models, applications and platforms it creates than for something more fundamental: the reorganization of economic activity around a new architecture.
The opportunity is enormous, but the outcome will be determined not just by technology architecture, but by the surrounding economic structures negotiated by those who control foundational rights. The opportunity for a Great Rearchitecture is to correct the foundational mistakes of the internet by redesigning the architecture of value creation itself. The lessons above highlight three key steps:
1. Disrupt Today’s Concentration.
The convergence of universal connectivity, software-defined assets and workflows, and context-capable AI creates the possibility of an open, infinitely composable ecosystem in which creators, developers, intermediaries, and end users can contribute to and capture value without surrendering control to a handful of dominant platforms. Agentic coding and orchestration can unbundle capabilities that have been locked inside proprietary silos and dynamically reassemble them into new packages of value, tailored to the needs of specific users and enterprises. Rather than requiring everyone to build on the same centralized platforms, this architecture can allow participants to combine resources, services, and intellectual property across an expanding network of specialized contributors. The result is a more fluid and competitive economy in which innovation can come from anywhere, and the ability to create value is no longer dependent on controlling the dominant distribution channel.
2. Restore Balance for Creators.
The second element of this rearchitecture is respect for strong intellectual property rights as a foundation for fair value allocation. In an economy where AI agents can rapidly generate and compose software, content, services, and physical-world capabilities, the ability to identify, protect, and compensate the creators of underlying innovations becomes essential. Strong patents and other enforceable IP rights can provide a mechanism for ensuring that the economic benefits generated by a composite system are allocated among the creators, developers, and platforms whose contributions make that system possible. Instead of treating foundational technology as a free resource to be appropriated by whoever controls the largest platform, the new architecture can recognize the independent value of each component and provide a basis for licensing, revenue sharing, and negotiated access. This does not require eliminating open collaboration or free software; it requires ensuring that openness is not synonymous with uncompensated appropriation, and that the ability to compose innovations does not extinguish the rights of those who created them.
3. Constrain Re-concentration.
The third element is the use of established legal and regulatory institutions to prevent the new architecture from reproducing the old concentration of power. We don’t necessarily need new laws; we do need to use institutions we inherited from an era where society had to deal with a similar level of disruption: the industrial revolution. The patent system can protect independent innovation and give creators a means to negotiate with powerful intermediaries, while antitrust law can constrain the ability of platforms to foreclose competition, acquire emerging rivals, or extend dominance from one layer of the ecosystem into another. Together, they can provide a counterweight to the natural tendency of successful networks to consolidate control.
The Great Rearchitecture is not simply a technological transformation; it is an opportunity to align technology, property rights, and market governance so that the benefits of AI-driven productivity are distributed more broadly. By combining infinitely composable architectures with enforceable rights and effective competition policy, the next generation of technology can break the recurring cycle in which innovation creates value, platforms capture it, and concentration ultimately limits the innovation that follows.
The goal of the next technological revolution should not merely be a more efficient version of the old economy. It should be an economy architected to distribute the benefits of technological abundance.