The Cooperative Default

In the [YOU] on AI Field Guide

The cycle that began with [YOU] on AI documents the AI transition from inside the experience of its most-affected communities: software developers, product builders, the people whose professional identities were structured by the very assumptions the transition is dissolving. What the cycle does not fully develop—and what the cooperative default supplies—is the empirical evidence that the dissolution of those structures is not simply a crisis. It is also a clearing.

Solnit's disaster research shows that when institutional structures collapse, the forms of cooperation that emerge are often more capable, more humane, and more equitable than the institutional structures that preceded them—and that this is not a coincidence but a consequence of the cooperative default being freed from the institutional arrangements that normally suppress it. The open-source AI ecosystem that has emerged alongside the commercial one—Llama, Mistral, the communities of researchers and engineers who share weights, code, and findings without institutional mediation—is a disaster community in this precise sense: spontaneous, cooperative, fragile, and more resilient per unit of resource than the commercially organized alternative.

The cooperative default also illuminates the specific character of the builder communities Segal describes: the three-in-the-morning building sessions, the public sharing of discoveries, the specific quality of generosity that characterizes the early-adopter communities around any powerful new tool. These are not personality traits. They are the cooperative default operating in a moment when the old institutional arrangements have not yet been reimposed and the new ones have not yet been built. The question Solnit presses—which the cycle leaves open—is whether this cooperative impulse can be institutionalized before the concentration of AI ownership reasserts the logic of extraction.

Origin

The concept is the central empirical finding of Solnit's A Paradise Built in Hell: The Extraordinary Communities That Arise in Disaster (2009), a book that assembled systematic evidence from five major disasters across a century of American and Mexican history. The research challenged a deeply held assumption in emergency management, political philosophy, and popular culture: the assumption that the removal of institutional order produces Hobbesian chaos, and that the primary task of emergency response is to prevent the population from preying on itself.

Solnit found the opposite. In every case she studied, the immediate post-disaster period was characterized by an extraordinary outpouring of mutual aid that the authorities were systematically unprepared to support and often actively suppressed. The elite panic—the conviction that institutional collapse produces looting and violence—was not merely wrong as a prediction. It was counterproductive as a response: the National Guard deployed to prevent imagined looting in post-Katrina New Orleans was preventing residents from sharing supplies from flooded stores, and the military response produced the disorder it was meant to prevent.

The concept has antecedents in Pyotr Kropotkin's Mutual Aid (1902), which documented cooperative behavior across species and human societies as evidence against Social Darwinist claims that competition was the primary mechanism of evolution and social life. Solnit's contribution is empirical specificity: rather than arguing from evolutionary theory about what humans should do, she documents what humans actually do when the institutional constraints that normally organize behavior are removed. The result supports Kropotkin's thesis but grounds it in concrete, recent, observable history rather than evolutionary conjecture.

Key Ideas

Cooperation Is Default, Not Designed. The disaster communities emerge without planning, without institutional support, and often against the active resistance of the authorities. This is the finding that makes the cooperative default genuinely surprising rather than merely heartwarming: cooperation is not produced by designing for it. It emerges when the institutional arrangements that normally suppress it are removed. The implication for the AI transition is that the cooperative impulse already present in the open-source communities and the knowledge-sharing networks does not need to be created. It needs to be institutionalized before the default arrangements of concentrated ownership suppress it.

Fragility and the Window. Disaster communities are temporary. They are typically displaced within weeks or months by the reassertion of normal institutional structures: government protocols, commercial contracts, proprietary claims on shared resources. The window during which the cooperative default is operating is real but short. Solnit's historical analysis of which cooperative impulses survived institutionalization and which did not reveals a consistent pattern: the ones that survived were the ones that found specific institutional vehicles—unions, laws, community land trusts, governance structures—before the window closed. The ones that did not find institutional expression were absorbed by the dominant arrangement or marginalized into irrelevance.

Elite Panic as Suppression Mechanism. The authorities' conviction that cooperation will give way to chaos is not merely wrong as a prediction. It functions as a suppression mechanism: by responding to imagined disorder with force, the authorities disrupt the cooperative networks that have spontaneously formed and produce the disorder they predicted. In the AI context, the regulatory frameworks emerging primarily from the AI companies themselves—designed to govern the technology while preserving the companies' structural advantages—may function similarly: genuine concern about AI risks combined with institutional arrangements that channel governance toward the priorities of concentrated ownership rather than the cooperative communities the technology has generated.