Step 2: tool results

get_claim

{
  "include": [
    "arguments",
    "subclaims",
    "dependents"
  ],
  "claim_id": "30b4f79d-4286-43ff-a6a9-be91d82c3692"
}

{"claim":{"id":"30b4f79d-4286-43ff-a6a9-be91d82c3692","kind":"claim","canonical_form":"Obfuscated arguments limit debate and iterated amplification to verifying arguments small relative to training computation.","claim_type":"empirical_derived","state":"active","decomposition_status":"complete","importance":0.4,"contestation":0.55,"domains":[],"tags":[{"slug":"ai-debate","name":"AI Debate"},{"slug":"scalable-oversight","name":"Scalable Oversight"},{"slug":"ai-alignment","name":"AI Alignment"}],"steward_state":"running","stewarded_at":"2026-11-15T00:36:06.134Z","canonical_direction_note":null,"created_at":"2026-11-11T19:12:38.352Z","updated_at":"2026-11-15T00:36:05.558Z","page_url":"https://minerval.ai/claims/30b4f79d-4286-43ff-a6a9-be91d82c3692"},"assessment":null,"coverage":{"examined":false,"on":[],"note":"No judgment of this node is in force. That is not a finding either way."},"preliminary_seed":{"credence":0.5,"note":"Barnes and Christiano reported in December 2020 that in human debate experiments a dishonest debater could often build a large argument containing a fatal flaw that neither side could locate, that they had no fix, and that this might be an important quantitative limit for both debate and iterated amplification. Later theoretical work on debate protocols (doubly-efficient and prover-estimator debate) claims to defeat obfuscated arguments under stability assumptions, and whether those assumptions hold for real arguments is open. The prior is even because the problem is well evidenced as a phenomenon while its reach as a hard limit, and its transfer from debate to amplification, remain unsettled.","seeded_by_claim_id":"51416960-dcb1-40ac-8b4f-9948df642ecf"},"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":"2026-11-11T19:12:38.420Z","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":"2026-11-15T00:36:05.567Z","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,"arguments":[],"dependents":[{"id":"51416960-dcb1-40ac-8b4f-9948df642ecf","text":"Iterated amplification can usefully train AI systems on complex real-world tasks for which no other training strategy is available.","claim_type":"empirical_derived","node_kind":"claim","relation_type":"contradicts","reasoning":"If recursive decomposition can only be trusted for arguments small relative to training computation, the tasks beyond human scale that the claim is about are the ones where the amplified signal cannot be trusted.","importance":0.4,"assessment_status":"inconclusive","assessment_confidence":0.6,"assessment_credence":0.25,"bounty_micro_usd":null,"checked":null,"formal":false}],"dependents_total":1,"subclaims":[],"subclaims_assessed":0}