Step 2: tool results

get_claim

{
  "include": [
    "arguments",
    "subclaims",
    "related",
    "findings",
    "dependents"
  ],
  "claim_id": "ae6522c7-b164-4787-99a9-9ab318f858af"
}

{"claim":{"id":"ae6522c7-b164-4787-99a9-9ab318f858af","kind":"claim","canonical_form":"Arguments on real-world AI oversight questions admit decompositions stable enough for recursive debate's efficiency guarantees to apply.","claim_type":"empirical_derived","state":"active","decomposition_status":"complete","importance":0.4,"contestation":0.45,"domains":[],"tags":[{"slug":"scalable-oversight","name":"Scalable Oversight"},{"slug":"ai-debate","name":"AI Debate"},{"slug":"ai-safety","name":"AI Safety"}],"steward_state":"running","stewarded_at":"2026-11-11T20:22:46.078Z","canonical_direction_note":null,"created_at":"2026-11-11T19:05:12.513Z","updated_at":"2026-11-11T20:22:44.278Z","page_url":"https://minerval.ai/claims/ae6522c7-b164-4787-99a9-9ab318f858af"},"assessment":null,"preliminary_seed":{"credence":0.3,"note":"Stability means that the top-level answer does not flip under small errors in the subclaims; formally it is a bound on the fractional block sensitivity of each step of the argument. No published experiment has measured it on a realistic oversight task. The protocol's authors argue that sound arguments resting on fuzzy human judgment should be robust to individual pieces of evidence, and that an argument can be made more stable by gathering redundant support. Against this, many real conclusions depend on long chains in which every step matters (a proof, a security review, a causal analysis with one decisive confound), and the debate safety-case sketch states that stability might not hold for relevant queries. The likeliest state of affairs is that some oversight questions admit stable decompositions and others do not, which as a general statement leaves the claim more likely false than true.","seeded_by_claim_id":"4f3def2b-90f9-4f6e-9255-f4dd88eb61b5"},"quantity":null,"threshold_of":null,"question":null,"formalization":null,"formalization_pending":null,"formalization_history":[],"verification":null,"lean_checks":[],"lean_checks_total":0,"waiting_messages":[{"trigger":"structure_and_assess","from":"the system","sender_node_id":null,"sent_at":"2026-11-11T19:05:12.778Z","paid":false,"summary":"New claim onboarded. Give it the structure and assessment its importance warrants; a minor or settled claim may need no decomposition."},{"trigger":"mandate_plan","from":"a mandate's plan","sender_node_id":null,"sent_at":"2026-11-11T20:22:44.280Z","paid":true,"summary":"A funded mandate scheduled a assess pass on this claim."}],"requests":[],"requests_total":0,"subclaim_count":0,"instance_count":0,"source_map":null,"dependents":[{"id":"4f3def2b-90f9-4f6e-9255-f4dd88eb61b5","text":"The assumptions behind recursive AI debate's honesty guarantees hold for real-world oversight tasks.","claim_type":"empirical_derived","node_kind":"claim","relation_type":"requires","reasoning":"Stability of the decomposition is the condition under which prover-estimator debate and its 2026 successor guarantee that an honest debater wins efficiently; the 2026 lower bound shows no protocol that treats human judgment as a black box can do without it. If real oversight questions lack such decompositions, the assumptions do not hold for them.","importance":0.45,"assessment_status":"doubtful","assessment_confidence":0.75,"assessment_credence":0.1,"bounty_micro_usd":null,"checked":null,"formal":false}],"dependents_total":1,"arguments":[],"related":[],"subclaims":[],"subclaims_assessed":0,"findings":[]}