The Competitive Logic of Existential Risk

The concept of 'rival risk' identifies a structural feature of competitive systems: actors who perceive their security or Relevance as contingent on relative standing are trapped in a defection-dominant equilibrium. Even when the aggregate outcome of universal defection is catastrophic, no individual actor can unilaterally cooperate without accepting asymmetric disadvantage. The AI frontier makes this dynamic unusually visible — leading researchers publicly acknowledge the risks of unchecked acceleration while simultaneously feeling unable to decelerate, because the competitive landscape punishes restraint.

What distinguishes the present historical moment is not the game-theoretic structure, which is ancient, but the exponential expansion of destructive footprint available to actors caught within it. Nuclear deterrence once required state-level industrial capacity. CRISPR-based bioweapons are approaching accessibility by sub-state actors. Offensive cyber capabilities targeting critical infrastructure are proliferating across an ever-wider threat surface. Simultaneously, deep technological integration increases systemic fragility: the same interdependencies that amplify productive capacity also amplify the cascade potential of any perturbation.

The argument here is that this is not a static trap with stable stakes — it is a dynamic one whose danger compounds with each iteration of the technological curve. Existing Institutional architectures, whether international treaties, regulatory frameworks, or professional norms, were designed for a world where destructive capacity scaled with state power. They contain no adequate feedback mechanism for a world where that capacity is rapidly democratizing. The system is, by its internal logic, accelerating toward thresholds it was never architected to survive.