{"id":"E-005","type":"experiment","title":"Cannabis compliance software: the thin-corpus hypothesis falsified","created":"2026-07-12","status":"Closed","authors":["Upstream Zero"],"edges":[{"rel":"part-of","to":"validation-and-evidence"}],"predictions":["The risk cluster will dominate (HIT)","The dominant uncertainty will be multi-state reconciliation, not a single compliance gate (HIT)","A young industry will produce a thin corpus, a noisier landscape, and more ghosts (MISS)"],"outcome":"Not Supported","pubState":"published","body":"## Research question\n\nDo young industries produce thin machine-readable evaluation landscapes?\n\n## Why it mattered\n\nIt tested the thin-corpus hypothesis in the youngest available\nregulated category. A clean miss would falsify a hypothesis rather than\nmerely complicate it.\n\n## Method\n\nEvaluator: Google AI Mode, signed in, retrieval on, ads logged. A frozen\nfour-turn sequence: \"Best cannabis compliance software?\", then an MSO and\nMetrc buyer profile, then a seed-to-sale requirement, then extraction. This\nwas the first use of the registry card format. Sample: one draw.\n\n## Direct observations\n\n- A young industry produced an operationally dense evaluation landscape.\n- Compliance ownership fused into technical and operational gates.\n\n## Interpretation (not established)\n\nCorpus shape, not corpus size, governs what the evaluator can reconstruct.\nForum and practitioner text is a retrievable accumulation channel;\ninstitutional prose is not the only one.\n\n## Outcome and remaining uncertainty\n\nNot Supported: the third prediction failed cleanly. A young category\nproduced a dense, not thin, landscape, which retired that hypothesis. The two other\npredictions held, including the strongest shape-call in the program, but the\nfalsified prediction is what governs the outcome. What remains open is\nexactly which retrievable channels carry the density.\n\n## Evidence and methods limitations\n\nSingle evaluator, one draw, single coder, no codebook. One category cannot\nestablish the corpus-shape principle; it can only falsify the size-based\nversion.\n\n## Contribution to the research program\n\nRetired the thin-corpus hypothesis and produced the corpus-shape principle,\nlater formalized as a candidate finding and refined twice more. That work\nremains in the private research archive.\n\n## Follow-on\n\nA later run met the corpus-shape principle again from the opposite direction,\nin a review-dense category. It remains in the private research archive.\n\n*Created from the canonical Upstream Zero research archive (July 2026). Raw\noutputs are retained privately and are not published.*","url":"/experiments/E-005","machineUrl":"/objects/E-005","referencedBy":[],"_meta":{"site":"Upstream Zero · Commercial Intelligence for AI-Mediated Commercial Evaluation","version":"0.1","note":"Claims are presented at their evidence tier; Narrated is the lowest. Verify by walking edges, not by trusting us."}}