Theory Audit — Field Note 05: Interface Race
Phase 2 — Theory Audit
Audit question
Is Interface Race a distinct AI-era competitive shift — or a current configuration of established platform economics, architectural control, complementary assets, default effects, dominant-design competition and regulatory pacing? If a residual remains, what exactly is new about an interface that mediates cognition and action rather than access alone?
39 papers across 6 folders
Folder A: Platform economics, gateways and network effects (6 papers). Folder B: Architecture, complementary assets and vertical integration (6 papers). Folder C: Interface effects, defaults, affordances and habit (6 papers). Folder D: Dominant designs, platform transitions and coexistence (6 papers). Folder E: AI assistants, devices and emerging gateway power (8 papers). Folder F: Governance pacing, regulation and contestability (7 papers).
Verdict: D — Observational concept retained, reclassified
Interface Race is not a newly discovered causal mechanism and should not be presented as a universal platform-transition law. It is a useful name for a current strategic configuration: competition to control the default cognitive-action gateway through which people discover, invoke, delegate to and remain connected with AI services. The residual is the emergence of a layer that combines information retrieval, framing, memory, identity, recommendation, service selection and delegated action — making distribution power and interpretive power coincide. Trigger case: Apple Inc. v. Liu (N.D. Cal., filed 10 July 2026).
Provisional definition
Competition among model providers, platform owners and device firms to become the default cognitive-action gateway through which users access, interpret and delegate to AI services. The condition becomes strategically significant when the interface can steer underlying service selection, retain context across interactions, and create switching or participation advantages that are not reducible to model quality alone.
Minimal mechanism chain (5 stages)
Interface placement → functional expansion → behavioural concentration → ecosystem reinforcement → capture risk. The sequence is cumulative but not automatic; any stage can stall, reverse or remain contestable.