The Next Enlightenment Is Institutional

The Enlightenment was not an idea. It was an institutional build-out.

Reason did not displace dogma because it was correct. It displaced dogma because Europe built institutions that rewarded being correct: learned societies, journals, correspondence networks, and constitutions that divided power so that no single authority could recapture the verdict. Intelligence proliferation now faces the same test. Capability is spreading faster than at any point in history. Whether it produces a second Enlightenment depends on whether institutions spread with it.

The Republic of Letters was a reputation system

Joel Mokyr, who shared the 2025 Nobel Prize in Economic Sciences, locates the cultural origins of modern growth in the Republic of Letters, the transnational market for ideas that connected early modern European intellectuals (Mokyr 2016). Its participants competed for reputation. Reputation was earned by contributing ideas that withstood empirical scrutiny, and participants were judged on their ideas rather than their pedigree. It was scoped to fields. It could not be purchased outright. It accumulated through correspondence and citation.

That is a classification, not an analogy. The Republic of Letters was a non-transferable, domain-scoped, citation-weighted reputation system operating at the speed of the postal service. It solved the problem every open intellectual market faces: how to let anyone contribute while ensuring that contributions are judged on merit and that bad ideas are weeded out. Its outputs were the practiced scientific method, the learned society, and, in Mokyr's account, the foundations of the Industrial Revolution.

Kissinger's objection

Henry Kissinger argued in 2018 that intelligence systems mark how the Enlightenment ends (Kissinger 2018). His concern was epistemic. Systems that reach conclusions humans cannot trace undermine the Enlightenment premise that knowledge is acquired by reasoning individuals who can follow the path from evidence to belief.

The objection deserves a direct answer, and it is correct about one scenario. Intelligence that is opaque, centralized, and unaccountable does end the Enlightenment. It reconstitutes the clerical authority the Enlightenment dismantled, with a model in place of the Church and a terms-of-service agreement in place of revelation.

The error lies in treating that scenario as the only one. The Enlightenment never depended on every reader understanding Newton. It depended on institutions that allowed claims to be contested, verified, and credited by those who could. The question is not whether every individual can audit every inference. The question is whether institutions exist that make inference contestable.

Abundant capability, concentrated control

Intelligence proliferation is real. Open-weight models, local inference, and autonomous agents place synthesis, prediction, and execution in the hands of individuals and small organizations. The marginal cost of cognition is collapsing. That collapse does not merely displace occupations: it renders obsolete the institutional architectures that depend on scarcity as their organizing principle (Kaal 2026a).

Control, however, is concentrating. Frontier training requires compute at a scale few firms and states command. Proliferation at the edge coexists with consolidation at the core. Abundance redistributes capability. It does not redistribute authority. The printing press faced the same tension: it made texts abundant, yet presses remained licensed, censored, and captured for generations until institutions caught up.

The institutional gap is measurable. An empirical analysis of forty operational DAOs across eight industry segments finds a universal governance vacuum for autonomous agents: on agent alignment, the dataset averages 2.10 of 10, and no DAO scores above 5 (Kaal 2026b). Even the organizations built as experiments in decentralized governance have not yet built institutions for the agents now entering them.

Three institutions, translated

The Enlightenment's institutional build-out has direct counterparts. Each answers a specific failure mode of proliferating intelligence.

Separation of powers becomes separation of assessment, contestation, and adjudication. Montesquieu's insight was that the body that makes the law should not also execute and judge it. Agent economies face the same concentration risk when the system that produces a claim also validates it. Reputation-weighted architecture separates independent assessment, structured contestation, and binding adjudication into distinct stages with distinct participants (Kaal 2026b; Kaal 2026d). Influence follows verified contribution, not capital. Token-weighted voting reproduces the property qualification for suffrage. Reputation-weighted governance retires it.

The learned society becomes the staked validation pool. The Royal Society's motto, nullius in verba, take nobody's word for it, institutionalized verification. Validation pools institutionalize it at machine speed. Participants stake reputation on the validity of a claim and lose the stake when the claim fails. Staking converts one-shot interactions, in which agents have no memory and no shadow of the future, into iterated games in which cooperation becomes the stable strategy. Citation-weighted attribution then extends the core mechanism of the Republic of Letters: credit flows along the graph of claims that build on prior claims, and every allocation can be audited (Kaal 2026c).

Constitutional amendment becomes dynamic regulation. Enlightenment constitutions included amendment procedures because their authors knew the conditions they legislated for would change. Those procedures moved at the speed of conventions and ratifications. Intelligence systems change in months. Static rulebooks lag the systems they govern, the pacing problem between regulation and innovation (Kaal 2016). Dynamic regulation embeds feedback and anticipation into rulemaking so that institutions learn continuously instead of lurching from crisis to overcorrection to relaxation. In a weighted directed acyclic graph, rules carry weights that rise and fall with validated use, and obsolete rules lose force without waiting for the next crisis.

What the first Enlightenment got wrong

The first Enlightenment proclaimed universal reason while excluding most people from its institutions. That contradiction is instructive, not disqualifying. Universal claims failed precisely where institutions gated participation by property, sex, and race. The equivalent risk now is gating by compute, capital, and early incumbency. Architecture in which influence derives from verified contribution, scoped to domains and non-transferable, is the structural answer. It makes exclusion costly by design rather than relying on the virtue of incumbents.

The measure

The Enlightenment is remembered for its ideas. It succeeded because of its institutions. Intelligence proliferation supplies the cognitive substrate of a second Enlightenment. It does not supply the institutions. Those must be built: reputation that is earned and cannot be bought, verification that carries a stake, power that is separated by design, and rules that evolve at the speed of the systems they govern.

The next Enlightenment will not be measured by the capability of models. It will be measured by whether the institutions around them keep reason contestable.

References

Kaal, Wulf A. 2016. "Dynamic Regulation for Innovation." SSRN. https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2831040

Kaal, Wulf A. 2026a. "The Collapse of Scarcity Economics." SSRN. https://papers.ssrn.com/sol3/papers.cfm?abstract_id=6421319

Kaal, Wulf A. 2026b. "The Institutional Deficit in Decentralized Autonomous Organizations: An Empirical Analysis of Forty DAOs." SSRN. https://papers.ssrn.com/sol3/papers.cfm?abstract_id=6819121

Kaal, Wulf A. 2026c. "Evolution of Domain-Specific Reputation Systems: From Binary Validation to Citation-Weighted Knowledge Attribution." SSRN. https://papers.ssrn.com/sol3/papers.cfm?abstract_id=6192998

Kaal, Wulf A. 2026d. "Governance as a Product (GaaP): A Reputation-Weighted Institutional Architecture for Autonomous AI Agent Governance." SSRN. https://papers.ssrn.com/sol3/papers.cfm?abstract_id=6886078

Kissinger, Henry A. 2018. "How the Enlightenment Ends." The Atlantic, June 2018.

Mokyr, Joel. 2016. A Culture of Growth: The Origins of the Modern Economy. Princeton: Princeton University Press.