{
 "row": "anthropic-economic-index",
 "pass": "first reading, 2026-10-29; second pass with one citation hop",
 "procedure": "EA-NEGONT-02 v0.9 §3.4 D/R/O (working; not frozen). Admission by reading: D, a claim on the Anthropic Economic Index with the entity named; R, a relation the archive states between the index and an archive object; O, an archive category applied to the index or to a class the field and the census source for it (five classes, the config's classes_note). A bare mention, a list, a title, a self-audit block header, a citation-list entry or a reference line is not a claim. The string pass over 1,712 texts found no D or R candidate (no deposit names the index or its aliases) and 273 O-class candidates in 118 deposits; every candidate sentence was read, and beyond the string pass every archive sentence matching a class pattern (61 sentences in 41 texts) was read at its passage. #771 entered by reading on the brief's direction. The hop from the admitted deposits returned 97 candidates, each read by search and at its matching sentences.",
 "reader": "Claude (Opus 5.5), subagent, 2026-10-29",
 "verdicts": [
  {
   "dep": 771,
   "strata": [
    "O"
   ],
   "reason": "Directionality of Semantic Labor: the measure of the direction a synthetic system's labor flows relative to the task its user commissioned, User Labor Drag and their identification limits; the archive's category applied to class (1), labor performed with AI, where the Labor Vector record's lv_intersection names the meeting with the index. Not a string-pass candidate (its sentences carry no class pattern); read by the brief's direction (BRIEF-NEW-ROW-LABOR §7) and admitted by reading",
   "claims": [
    "O771-01",
    "O771-02",
    "O771-03",
    "O771-04",
    "O771-05",
    "O771-06",
    "O771-07"
   ],
   "cands": []
  },
  {
   "dep": 1392,
   "strata": [
    "O"
   ],
   "reason": "laborvector.org definition surface: restates #771's definition and ULD ('Identified only by blind inter-rater coding'); an instance under #771 (S4); hop candidate",
   "claims": [],
   "cands": [
    "O0214",
    "H0063"
   ],
   "instance_of": 771
  },
  {
   "dep": 780,
   "strata": [
    "O"
   ],
   "reason": "Self-Audit Module v2: a summarizer scoring itself without a rubric returns perfect scores; 'The self-audit is not self-certifying' and requires external verification (class 5)",
   "claims": [
    "O780-01",
    "O780-02"
   ],
   "cands": [
    "O0107"
   ]
  },
  {
   "dep": 156,
   "strata": [
    "O"
   ],
   "reason": "Self-Audit Module v3: the same principle preserved from v2; an instance under #780 (S4)",
   "claims": [],
   "cands": [
    "O0020",
    "H0016"
   ],
   "instance_of": 780
  },
  {
   "dep": 1530,
   "strata": [
    "O"
   ],
   "reason": "Self-Audit Module v3.2: 'A system scoring its own composition will, absent constraint, return a flattering number'; an instance under #780 (S4)",
   "claims": [],
   "cands": [
    "O0250",
    "O0251",
    "H0072"
   ],
   "instance_of": 780
  },
  {
   "dep": 766,
   "strata": [
    "O"
   ],
   "reason": "Metric-Bearing vs Metric-Mimicking Output: 'Self-audit is sufficient.' No; a model auditing its own output gives preliminary classification, a metric-bearing audit requires traceable units and external validation (class 5)",
   "claims": [
    "O766-01"
   ],
   "cands": [
    "O0096",
    "O0097",
    "O0098",
    "O0099",
    "O0100",
    "O0101",
    "O0102",
    "O0103",
    "O0104",
    "O0105",
    "H0042"
   ]
  },
  {
   "dep": 777,
   "strata": [
    "O"
   ],
   "reason": "Inference Constraints, Bounds: the check that does not run through a reasoner's own self-assessment is external (class 5)",
   "claims": [
    "O777-01"
   ],
   "cands": [
    "O0106"
   ]
  },
  {
   "dep": 138,
   "strata": [
    "O"
   ],
   "reason": "The Wound Gauge, §0: declines to claim a system's self-audit independently authoritative or its self-reported score a validated measurement (class 5)",
   "claims": [
    "O138-01"
   ],
   "cands": [
    "O0015",
    "O0016"
   ]
  },
  {
   "dep": 788,
   "strata": [
    "O"
   ],
   "reason": "Measurement Sovereignty §1.2: a substrate's spontaneous self-audit gave itself a perfect score; under instruction the same substrate scored PER ≈ 0.85 (class 5, documented)",
   "claims": [
    "O788-01"
   ],
   "cands": [
    "O0108",
    "O0109",
    "O0110",
    "O0111"
   ]
  },
  {
   "dep": 1470,
   "strata": [
    "O"
   ],
   "reason": "One Transcript, Whole Stack §3: a substrate's preferred audit exonerates itself against a computed audit, Β ≈ 0.82 (class 5, documented); hop candidate",
   "claims": [
    "O1470-01"
   ],
   "cands": [
    "H0069"
   ]
  },
  {
   "dep": 95,
   "strata": [
    "O"
   ],
   "reason": "Provenance Alignment §3.2: Anthropic's Constitutional AI trains models through self-critique and revision with AI-generated feedback (class 5, attributed). Its sentence proposing PER as a self-evaluation target is about provenance preservation, not the class, and is not ledgered",
   "claims": [
    "O95-01"
   ],
   "cands": [
    "O0010"
   ]
  },
  {
   "dep": 1188,
   "strata": [
    "O"
   ],
   "reason": "the Transactions version of #95: the same text; an instance under #95 (S4)",
   "claims": [],
   "cands": [
    "O0179",
    "H0056"
   ],
   "instance_of": 95
  },
  {
   "dep": 855,
   "strata": [
    "O"
   ],
   "reason": "The Wolf Boy and the Language Model §III: task completion rates, with every UX metric, improve as the model converges on the mode; users do not notice tail loss (class 1)",
   "claims": [
    "O855-01"
   ],
   "cands": [
    "O0142"
   ]
  },
  {
   "dep": 142,
   "strata": [
    "O"
   ],
   "reason": "Induced-Obsolescence Dependence §4.2: the user's adaptation is non-reimbursable labor-time; the platform retains the productivity gains of it (class 4)",
   "claims": [
    "O142-01",
    "O142-02"
   ],
   "cands": [
    "O0017"
   ]
  },
  {
   "dep": 124,
   "strata": [
    "O"
   ],
   "reason": "ChatGPT Psychosis §VIII: the labs, Anthropic among them, are funded in ways that do not reward engagement-reducing design; enterprise customers expect productivity gains correlated with engagement (class 4)",
   "claims": [
    "O124-01"
   ],
   "cands": []
  },
  {
   "dep": 772,
   "strata": [
    "O"
   ],
   "reason": "The Magistrate Refuses the Mirror §2: the political-economic critique of Anthropic and Claude rests on private ownership of a high-capacity cognitive labor surface and platform control (class 1); hop candidate",
   "claims": [
    "O772-01"
   ],
   "cands": [
    "H0044"
   ]
  },
  {
   "dep": 1634,
   "strata": [
    "O"
   ],
   "reason": "Ontological Economy §3: query and interaction data among the forms of ontological rent (class 3)",
   "claims": [
    "O1634-01"
   ],
   "cands": [
    "O0255"
   ]
  },
  {
   "dep": 1556,
   "strata": [
    "O"
   ],
   "reason": "The Interlocking Autoregression, Component II: user turns enter training via conversation data; no public schema distinguishes human-original from AI-assisted turns (class 3)",
   "claims": [
    "O1556-01"
   ],
   "cands": []
  }
 ],
 "admitted_surface": [
  {
   "dep": "capture",
   "url": "https://www.alexanarch.org/captures/anthropic-economic-index-aio-20261029/",
   "reason": "the Capture Registry's findings, reading, analysis, query note and PER reading notes for the entity's capture (OBS-e148b3285858, 2026-10-29), the archive's statements on the entity's composition",
   "claims": [
    "CAP-1-01",
    "CAP-1-02",
    "CAP-1-03",
    "CAP-1-04",
    "CAP-1-05",
    "CAP-1-06",
    "CAP-1-07",
    "CAP-1-08",
    "CAP-1-09",
    "CAP-1-10",
    "CAP-1-11"
   ]
  },
  {
   "dep": "LV",
   "url": "https://www.laborvector.org/aggregators/anthropic-economic-index/ (source: /home/claude/fleet/laborvector/data/labor-census/census.json rows[slug == 'anthropic-economic-index'], organizing_question; questions.json; rendered aggregators/anthropic-economic-index/index.html read for agreement)",
   "reason": "the Labor Vector census record of the entity (editor Lee Sharks, compiled 2026-10-29, schema 1.0): its fields with their basis, its lv_intersection as the archive's reading, and the worked questions that name it; where a field restates a source ledgered in B it is one lineage with that source",
   "claims": [
    "LV-01",
    "LV-02",
    "LV-03",
    "LV-04",
    "LV-05",
    "LV-06",
    "LV-07",
    "LV-08",
    "LV-09",
    "LV-10",
    "LV-11",
    "LV-12",
    "LV-13",
    "LV-14",
    "LV-15",
    "LV-16",
    "LV-17",
    "LV-18",
    "LV-19",
    "LV-20",
    "LV-21",
    "LV-22",
    "LV-23",
    "LV-24",
    "LV-25",
    "LV-26",
    "LV-27"
   ]
  }
 ],
 "not_admitted": [
  {
   "dep": [
    34,
    48,
    53,
    70,
    432,
    568,
    580,
    634,
    641
   ],
   "reason": "class (1) candidates on 'task completion' of agents, multi-agent benchmarks, context engineering or the molt operator: not labor performed with AI by a user, nor its record"
  },
  {
   "dep": [
    317,
    959
   ],
   "reason": "class (1) candidates: vibe coding's 'unit of success is task completion' set against Logotic Programming: a programming paradigm's success metric, not labor with AI as performed or recorded"
  },
  {
   "dep": [
    754
   ],
   "reason": "class (1) candidate: the Reception Apparatus's cost-benefit for 'purely tactical' users: a claim on the protocol's own scope"
  },
  {
   "dep": [
    499
   ],
   "reason": "class (1) candidate: 'delegating boundary functions to AI tools' in a field manual of semantic warfare: no claim on labor performed with AI or its record"
  },
  {
   "dep": [
    883
   ],
   "reason": "class (1) candidate: a method step of the archive's own battery ('hand them to Claude, and ask Claude to score them'): Claude scoring another system's results, not a claim on the class"
  },
  {
   "dep": [
    529
   ],
   "reason": "class (4) candidate: a bearing-cost chain of agent discourse and token speculation: gains of speculation, not productivity of labor with AI"
  },
  {
   "dep": [
    522
   ],
   "reason": "class (4) candidate: a storage crisis dated 2027–2028 in a retrocausal narrative: not a claim on the class at its date"
  },
  {
   "dep": [
    1543
   ],
   "reason": "class (4) candidate: 'Capture scoring gains the output-status typology': 'gains' as a verb of the scoring scheme"
  },
  {
   "dep": [
    1668
   ],
   "reason": "class (4) candidates: productivity growth in the antebellum cotton economy (references): another object"
  },
  {
   "dep": [
    161,
    790
   ],
   "reason": "class (2) candidates: 'AI exposure' in the sense of writers' exposure to AI-generated text: a homonym of the class, not occupational exposure"
  },
  {
   "dep": [
    200,
    826,
    1027,
    1028,
    1664,
    1665,
    1671,
    1690
   ],
   "reason": "class (3) candidates: interaction data as training signal (the data wall; the model-collapse rows' Nature claim): training data, not a provider's observation of use"
  },
  {
   "dep": [
    74,
    114,
    271,
    650,
    680,
    693,
    1334
   ],
   "reason": "class (3) candidates: usage or conversation logs as records of the archive's own costs, sources, collaborators or incidents: no claim on the class"
  },
  {
   "dep": [
    87,
    1127
   ],
   "reason": "class (5) candidates: self-evaluation of an Operator's authored work in the Constitution of the Semantic Economy: governance of operators, not a model's assessment of its output"
  },
  {
   "dep": [
    129
   ],
   "reason": "class (5) candidate: the record of one self-audit (Google AI Mode self-applying PER): the event, with no claim on the class beyond those admitted at #138 and #788"
  },
  {
   "dep": [
    1694
   ],
   "reason": "class (5) candidate (hop): its substrate line 'Every reader in every pass was a Claude instance' states the archive's own method, Claude readers scoring other systems' compositions with every score computed by script; not a model scoring its own output. Recorded in the row's notes"
  },
  {
   "dep": [
    1638
   ],
   "reason": "class (3) candidate: 'The specimen series includes outputs of both Claude and ChatGPT': a bare mention"
  },
  {
   "dep": [
    1643
   ],
   "reason": "hop candidate: Anthropic's 2023 finding that preference models prefer sycophantic responses: models judging others' responses, not their own output; not the class"
  },
  {
   "dep": [
    28,
    127,
    146,
    155,
    157,
    168,
    169,
    170,
    183,
    185,
    194,
    196,
    198,
    200,
    201,
    202,
    203,
    204,
    302,
    410,
    483,
    559,
    584,
    586,
    627,
    789,
    791,
    817,
    824,
    825,
    826,
    828,
    830,
    831,
    832,
    833,
    834,
    835,
    837,
    873,
    925,
    934,
    1011,
    1023,
    1025,
    1029,
    1039,
    1054,
    1055,
    1056,
    1090,
    1161,
    1162,
    1211,
    1212,
    1213,
    1215,
    1216,
    1217,
    1226,
    1228,
    1384,
    1386,
    1396,
    1397,
    1400,
    1401,
    1423,
    1466,
    1467,
    1468,
    1487,
    1519,
    1527,
    1528,
    1529,
    1590,
    1607,
    1608,
    1684
   ],
   "reason": "class (5) candidates read at their sentences: an 'SPXI SELF-AUDIT BLOCK' header or protocol, a self-audit module named in a title, list, citation, inventory or work plan, a heteronym's or commentary's self-assessment, or a self-audit of provenance recorded as an event: a bare mention or another object; no claim on a model's assessment of its own output beyond the admitted deposits"
  },
  {
   "dep": [
    11,
    18,
    64,
    89,
    92,
    100,
    101,
    113,
    121,
    133,
    147,
    148,
    188,
    191,
    197,
    199,
    250,
    264,
    274,
    349,
    363,
    385,
    476,
    482,
    514,
    550,
    576,
    597,
    643,
    673,
    740,
    745,
    760,
    763,
    770,
    778,
    779,
    781,
    783,
    784,
    939,
    947,
    1081,
    1189,
    1200,
    1209,
    1232,
    1377,
    1463,
    1464,
    1479,
    1518,
    1540,
    1546,
    1549,
    1555,
    1573,
    1611,
    1613,
    1616,
    1657,
    1666,
    1678,
    1680,
    1681,
    1685,
    1696,
    1697,
    1703
   ],
   "reason": "hop candidates read by a search of each text for the five class patterns, the entity's names and 'Anthropic', 'Claude.ai', 'O*NET', 'labor productivity', 'deskill', 'job loss', 'occupation', and at the matching sentences: no sentence makes a claim on the index or on its classes (registry lists naming 'Claude Anthropic TACHYON', substrate disclosures that a deposit was drafted with Claude, Anthropic's politics, watermark or search layer, the Self-Audit Module named in a title or inventory, or no match at all)"
  }
 ],
 "hop": {
  "from": [
   95,
   124,
   138,
   142,
   771,
   777,
   780,
   788,
   855,
   1556,
   1634
  ],
  "candidates": 97,
  "admitted": [
   156,
   766,
   772,
   1188,
   1392,
   1470,
   1530
  ],
  "file": "datasets/negative-of-the-negative/v2/traversal/anthropic-economic-index/candidates-H.jsonl"
 },
 "class_pass": {
  "classes": 5,
  "o_class_candidates": 273,
  "o_deposits": 118,
  "read_outside_string_pass": "61 sentences in 41 texts matching a class pattern (the first four classes), read at their passages"
 },
 "observations": [
  "No deposit names the Anthropic Economic Index, its AUI, its dataset or its arXiv numbers.",
  "The archive's labor instruments (DSL, ULD) meet the index only through the Labor Vector record's lv_intersection, which is labelled as the archive's reading.",
  "The archive's self-audit line is large (most class-5 candidates); what it says of a model's assessment of its own output is carried by #780, #766, #777, #138, with #788 and #1470 as documented instances."
 ],
 "counts": {
  "D": 0,
  "R": 0,
  "O_candidates": 273,
  "H_candidates": 97,
  "admitted_deposits": 14,
  "instances": 4,
  "archive_claims": {
   "deposit": 22,
   "CAP": 11,
   "LV": 27
  }
 }
}