The World Game

In The You On AI Field Guide

Fuller proposed the World Game at Southern Illinois University in 1961 and refined it through the next two decades. Partial implementations were attempted — student exercises at SIU, workshops hosted by the Buckminster Fuller Institute, various pedagogical adaptations — but the full-fidelity, real-time, globally comprehensive simulation Fuller specified remained computationally out of reach. Modeling cascading effects across interconnected systems — energy policy in one region propagating through agriculture, water, economics, population movement, and ecology elsewhere — required processing power that did not exist in any decade during which Fuller continued to advocate for the concept.

The political barrier was at least as formidable. Fuller argued that demonstration of feasibility would generate the political will to implement it — that showing, with the rigor of military simulation, that comprehensive prosperity was achievable would erode resistance to implementation. The logic was sound within his framework: if the problem is ignorance of what is possible, the solution is demonstration of what is possible. But partial implementations over subsequent decades suggested a complication Fuller did not fully accommodate. Climate models, epidemiological simulations, and economic projections are limited-domain versions of the concept. Each has produced actionable insights; each has demonstrated that computational proof of feasibility does not automatically generate political will.

The inference is not that simulation is useless. It is that Fuller's theory of change underestimated the structural resistance comprehensive solutions encounter from interests that benefit from the competitive status quo. The Great Pirates do not oppose comprehensive solutions because they are ignorant of feasibility; they oppose them because comprehensive solutions redistribute advantages the competitive system concentrates in their favor.

AI changes the computational dimension decisively. A sufficiently capable system could model global resource flows, ecological parameters, economic interactions, and population dynamics at fidelity that would have seemed fantastical even a decade ago. The political dimension may shift through a mechanism previous technologies could not access: specificity. Previous demonstrations operated at a generality that political discourse could absorb and neutralize. A contemporary World Game could produce specific intervention designs, for specific regions, with specific projected outcomes, at specific confidence levels. The city planner in Nairobi could test a water infrastructure proposal against the global model and trace cascading effects. Whether specificity is sufficient to overcome resistance that generality could not overcome is an open question — perhaps the most consequential one in the application of AI to planetary governance.

Origin

Fuller first proposed the World Game at the World Design Science Decade initiative in 1961 and the Southern Illinois University symposia that followed. He continued to refine and promote the concept until his death in 1983.

Alex Soojung-Kim Pang

Josh Pang's 2018 UC Santa Cruz thesis reframed the World Game as a computational system — a synthesis of database, projection, and machine learning simulation — arguing that the formal structure Fuller envisioned was technically achievable for the first time.

Key Ideas

Design-science alternative to war games. Same computational rigor, redirected from competitive destruction to comprehensive prosperity.

The calculation, not the aspiration. Fuller insisted Earth's resources were sufficient for all passengers; the barrier was organizational, not technical.

The computational barrier has fallen. What was fantastical in 1961 is now within reach. The processing power exists. The tool is finally adequate to the game.

Specificity as political lever. AI-powered simulation may produce demonstrations specific enough to resist the generalizing dismissals political discourse uses against comprehensive proposals.

The governance question remains. The game cannot be a product of any corporation or single government. It requires a governance structure representing all passengers — and that structure does not yet exist.