The Two-Clock Model: Structural Presence and AI Perception of Technology Entities

Frozen dataset and analysis code for the Two-Clock Model, which separates an entity's structural presence S(t) — how machine-readable its website is — from AI perception P(t) — what AI platforms actually know and cite about it. The perception gap G(t) = S(t) − P(t) is tested on a panel of 50 technology entities launched Jan 2023–Mar 2025, using a model-cutoff natural experiment over dated OpenAI snapshots (P(t)), archived homepage snapshots (S(t)), and weekly GDELT news mentions (C(t)). Includes the entity roster, three harvester scripts, and a citation-precision audit. All data frozen as v1 on 2026-07-08. Note: P(t) is a single-run pilot; citation precision is bimodal — use the PASS verdicts in ct_artlist_precision.csv for high-confidence citation series. Full methodology and limitations in README.md.