What the AI Race Has Given Us and What It Requires Next

Counts differ because each database indexes a different set of publications. We treat OpenAlex as the canonical count; Google Scholar is not shown (no API, and crawling it violates its ToS).

The AI race is often framed as a choice between unrestricted acceleration and a pause in development. This article reframes the issue as one of governing competition in a commercially valuable, geopolitically salient, and increasingly infrastructural field. Drawing on scholarship on responsible innovation, foundation models, HCI, organizational governance, and political economy, it argues that AI progress is cumulative and that competition can accelerate and diffuse capabilities and governance practices. The central problem is not competition itself, but competition without institutional steering. The article develops “race logic” as a middle-range causal mechanism linking competitive pressure to organizational choices, interface design, deployment, user behavior, and social outcomes. It proposes shared evaluation and incident infrastructure and enforceable lifecycle release gates as foundational interventions, supported by staged openness, participation, distributive accountability, international coordination, auditing, procurement, compute oversight, independent review, and anti-capture safeguards.