{"schema_version":1,"experiment_id":"eoa_inverse_innovation_exp03_full320_20260801","cell_id":"computability_boundary_mapping__communication_media_studies","trajectory_id":"R","attempt_index":0,"archetype_slug":"computability_boundary_mapping","domain_slug":"communication_media_studies","decision":"CANDIDATE","problem_id":"universal_interactive_media_policy_decision","causal_lever_id":"scope_typed_multistate_moderation_routing","proposal":{"problem":"Platforms and news organizations may require an automated moderator to return an exact, terminating policy-compliance verdict for every digital-media artifact, even when the admitted class includes arbitrary scripts, interactive behavior, generated continuations, or links whose behavior unfolds during execution. Timeouts, unsupported inputs, and inconclusive searches are then forced into allow/block decisions. HYPOTHESIS: for nontrivial behavioral policies over unrestricted programmable media, the total-exact requirement can include undecidable instances.","actors_substrate":["platform policy owners","moderation engineers and vendors","newsroom verification teams","content creators and publishers","audiences and people represented in media","appeals reviewers and regulators","programmable and interactive media artifacts"],"observable_state":"An unrestricted universal guarantee is advertised, while production logs show timeouts or unsupported artifacts converted into Boolean verdicts, undocumented scope narrowing, and no distinction among compliant, violating, unknown, out-of-scope, and system failure.","consequence":"Potentially unsafe artifacts receive false clearance, lawful expression is falsely blocked, appeals cannot reconstruct the operative guarantee, and engineering effort is spent pursuing total coverage that may be impossible under the declared class.","affected_objective":"Truthful, consistent, reviewable moderation with bounded false-clearance and false-blocking risk.","structural_mapping":[{"archetype_element":"Open-ended class plus universal exact terminating requirement","domain_realization":"Every admitted media artifact must receive a correct allow/block verdict under an expressive behavioral policy.","claim_kind":"HYPOTHESIS"},{"archetype_element":"Explicit representation and computation model","domain_realization":"Specify artifact formats, embedded code, external retrieval, interaction, randomness, and permitted analyst or service calls.","claim_kind":"INFERENCE"},{"archetype_element":"Constructive and impossibility evidence in parallel","domain_realization":"Seek a total checker for restricted media fragments while testing whether unrestricted behavioral compliance admits a valid halting reduction.","claim_kind":"HYPOTHESIS"},{"archetype_element":"Decidable regions and weaker honest fallback","domain_realization":"Enforce checkable fragments and route other artifacts to bounded analysis, sound one-sided findings, explicit unknown, or accountable review.","claim_kind":"INFERENCE"},{"archetype_element":"Versioned guarantee and reclassification","domain_realization":"Link each public moderation claim to its policy language, artifact model, evidence, fallbacks, and recheck triggers.","claim_kind":"INFERENCE"}],"component_map":[{"component":"Problem-Class Specification","status":"direct","domain_realization":"Define the admitted artifacts and policy predicates."},{"component":"Instance Representation Contract","status":"direct","domain_realization":"Fix encodings for static, executable, linked, and interactive media."},{"component":"Computation Model Contract","status":"direct","domain_realization":"Declare execution, network, oracle, randomness, and human capabilities."},{"component":"Solvability Guarantee Profile","status":"direct","domain_realization":"Separate exact decision, one-sided detection, bounded findings, and review."},{"component":"Quantifier and Scope Map","status":"direct","domain_realization":"Expose whether claims cover every artifact, a fragment, or sampled instances."},{"component":"Computability Status Lattice","status":"adapted","domain_realization":"Classify policy-artifact pairs as decidable, recognizable, relative, undecidable, or unresolved."},{"component":"Constructive Procedure Witness","status":"direct","domain_realization":"Provide a checker and termination/correctness proof for any claimed decidable fragment."},{"component":"Reduction Preservation Contract","status":"direct","domain_realization":"Require a computable behavior-preserving map from halting instances to moderation instances."},{"component":"Computability Impossibility Certificate","status":"direct","domain_realization":"Retain the checked reduction, if established, for the exact unrestricted claim."},{"component":"Assumption Register","status":"direct","domain_realization":"Record policy semantics, encodings, environmental access, and promise assumptions."},{"component":"Decidable Subclass Map","status":"adapted","domain_realization":"Map enforceable static-media and restricted-script fragments."},{"component":"One-Sided Recognition Contract","status":"adapted","domain_realization":"Permit witnessed violation findings without claiming complete clearance."},{"component":"Unknown and Nontermination Policy","status":"direct","domain_realization":"Keep unknown, timeout, out-of-scope, failure, allow, and block distinct."},{"component":"Fallback Solution Contract","status":"direct","domain_realization":"Specify guarantees for bounded checks, abstractions, and human escalation."},{"component":"Computability Guarantee Record","status":"direct","domain_realization":"Version the shipped claim and supporting evidence."},{"component":"Recheck Trigger","status":"direct","domain_realization":"Reassess after policy-language, media-format, model, or oracle changes."},{"component":"Termination Condition","status":"direct","domain_realization":"Set explicit budgets for bounded modes and guaranteed termination within fragments."},{"component":"Scope Boundary","status":"direct","domain_realization":"Mechanically admit, reject, or quarantine artifacts by supported class."},{"component":"Decision Record","status":"direct","domain_realization":"Record routing, verdict type, operative guarantee, and reason."},{"component":"Uncertainty Residue","status":"direct","domain_realization":"Preserve open proof obligations and inconclusive artifact findings."},{"component":"Independent Proof Review","status":"direct","domain_realization":"Have an independent reviewer check theorem fit and reduction correctness."},{"component":"Complexity Follow-On Gate","status":"direct","domain_realization":"Assess cost only after establishing in-principle decidability."}],"mechanism_dispositions":[{"slug":"abstract_interpretation_or_model_checking","disposition":"selected_supporting","contribution_type":"OPERATIONAL","adaptation_or_rejection":"Use sound finite abstractions for restricted programmable media; label false-alarm risk.","counterfactual_removal":"Fallback loses a potentially sound clearance mode but the boundary remains."},{"slug":"bounded_domain_exhaustive_search","disposition":"selected_supporting","contribution_type":"TEST_DESIGN","adaptation_or_rejection":"Exhaust finite artifact/state/depth bounds without extrapolation.","counterfactual_removal":"The pilot loses complete within-bound checks."},{"slug":"computability_boundary_decision_record","disposition":"selected_load_bearing","contribution_type":"SAFETY_GUARDRAIL","adaptation_or_rejection":"Version scope, shipped guarantee, evidence, and triggers.","counterfactual_removal":"Guarantees can drift silently and outputs become unauditable."},{"slug":"computational_complexity_analysis","disposition":"selected_supporting","contribution_type":"OPERATIONAL","adaptation_or_rejection":"Apply after a fragment is shown decidable.","counterfactual_removal":"Computable fragments may still be operationally infeasible."},{"slug":"constructive_algorithm_and_correctness_proof","disposition":"selected_load_bearing","contribution_type":"CORE_CAUSAL","adaptation_or_rejection":"Required before promising total exact decisions on a fragment.","counterfactual_removal":"Positive decidability claims rest on tests rather than class-wide evidence."},{"slug":"diagonalization_impossibility_proof","disposition":"considered_rejected","contribution_type":"NONE","adaptation_or_rejection":"A direct self-reference proof is unnecessary if a cleaner domain reduction exists.","counterfactual_removal":"No change; reduction evidence remains available."},{"slug":"enumeration_and_dovetailing","disposition":"considered_rejected","contribution_type":"NONE","adaptation_or_rejection":"Unbounded operation is unsuitable for live moderation; use bounded explicit-unknown protocol.","counterfactual_removal":"No material change."},{"slug":"fallback_mode_router","disposition":"selected_load_bearing","contribution_type":"OPERATIONAL","adaptation_or_rejection":"Route by enforced class to exact, abstract, bounded, unknown, or review modes.","counterfactual_removal":"Weaker guarantees are likely to be laundered into Boolean verdicts."},{"slug":"halting_problem_reduction","disposition":"selected_load_bearing","contribution_type":"CORE_CAUSAL","adaptation_or_rejection":"Test unrestricted behavioral-policy decision by embedding program termination while preserving the policy answer.","counterfactual_removal":"The claimed impossibility would lack decisive evidence."},{"slug":"language_fragment_restriction","disposition":"selected_load_bearing","contribution_type":"CORE_CAUSAL","adaptation_or_rejection":"Define mechanically enforceable static or restricted-script languages.","counterfactual_removal":"No dependable total-answer region is recovered."},{"slug":"many_one_reduction_proof","disposition":"selected_supporting","contribution_type":"TEST_DESIGN","adaptation_or_rejection":"Formalize totality, computability, and answer preservation of the halting embedding.","counterfactual_removal":"The impossibility argument becomes vulnerable to an invalid transfer."},{"slug":"promise_problem_restriction","disposition":"considered_rejected","contribution_type":"NONE","adaptation_or_rejection":"Unenforced promises permit authoritative behavior outside the guarantee; prefer syntactic admission.","counterfactual_removal":"No change to the enforceable-fragment design."},{"slug":"proof_by_counterexample","disposition":"selected_supporting","contribution_type":"TEST_DESIGN","adaptation_or_rejection":"Use legitimate in-scope artifacts to refute overbroad vendor guarantees.","counterfactual_removal":"Overclaims become harder to disconfirm cheaply."},{"slug":"proof_checking","disposition":"selected_load_bearing","contribution_type":"SAFETY_GUARDRAIL","adaptation_or_rejection":"Independently check constructive and reduction proofs against the stated media model.","counterfactual_removal":"A subtle proof or formalization gap could govern live moderation."},{"slug":"reduction_direction_checklist","disposition":"selected_supporting","contribution_type":"SAFETY_GUARDRAIL","adaptation_or_rejection":"Gate any hardness claim on source-to-target direction and declared assumptions.","counterfactual_removal":"A reversed or incomplete reduction may be accepted."},{"slug":"semi_decision_with_explicit_unknown","disposition":"selected_load_bearing","contribution_type":"OPERATIONAL","adaptation_or_rejection":"Return witnessed violation or UNKNOWN at a declared bound, never fabricate compliance.","counterfactual_removal":"Timeouts again collapse into unsafe allow/block verdicts."},{"slug":"theorem_prover_guided_search","disposition":"considered_rejected","contribution_type":"NONE","adaptation_or_rejection":"Optional proof discovery adds no essential guarantee beyond independent certificate checking.","counterfactual_removal":"Proof discovery may be slower, but the design is unchanged."},{"slug":"turing_reduction_analysis","disposition":"considered_rejected","contribution_type":"NONE","adaptation_or_rejection":"Relative degrees are unnecessary for the initial total-decider boundary and obscure one-sided guarantees.","counterfactual_removal":"No material change."}],"causal_chain":["Make artifact class, policy predicate, model, quantifiers, and output guarantee explicit.","Test unrestricted behavioral compliance using a checked source-to-target reduction while constructing total checkers for enforceable fragments.","Restrict exact decisions to fragments with constructive termination and correctness evidence.","Route remaining inputs to labeled abstraction, bounded search, explicit unknown, or accountable review.","Preserve distinct statuses and versioned records, preventing timeout or weak evidence from becoming false certainty."],"baseline":"A monolithic classifier returns allow/block for all submitted artifacts; timeouts and unsupported formats are handled by undocumented defaults, and benchmark accuracy substitutes for a class-wide guarantee.","nearest_rival":"A better-trained probabilistic moderation classifier with confidence thresholds and human review, but without proving which inputs admit total exact decisions or preserving model-relative unknown states.","authority_safety":{"affected_parties":["creators and publishers","audiences","people depicted or discussed","moderators and appeals staff","platform and newsroom operators"],"decision_authority":"The platform or newsroom policy authority owns scope and deployment; independent proof reviewers approve boundary evidence, and appeals personnel retain authority over consequential individual cases.","authorized_first_step":"Run a sandboxed, non-enforcement pilot on one versioned policy and a finite synthetic corpus containing static items, restricted scripts, and adversarial programmable artifacts; measure routing coverage, status preservation, false-clearance/blocking, review load, and proof obligations.","excluded_actions":["automatic removal, account sanction, or publication denial during the pilot","claiming undecidability from timeout or failed proof search","generalizing bounded results beyond their bound","treating UNKNOWN as compliant or violating","executing untrusted media outside a sandbox"],"halt_rollback":"Halt if statuses collapse downstream, the admission check can be bypassed, abstraction soundness is unsupported, the reduction fails review, or pilot errors exceed predeclared limits; revert to existing human-reviewed handling and invalidate the boundary record."}},"negative_tests":{"strongest_counterevidence":"The proposed unrestricted class may not actually include arbitrary executable behavior, or its policy may reduce to a known total syntactic check. A verified total decider for the full deployed representation would defeat the hypothesized impossibility boundary.","analogy_break":"Media truth, harm, and meaning may depend on ambiguous interpretation, future social response, or unavailable empirical facts. Computability analysis cannot turn an unstable construct into a faithful predicate and may prove a theorem about the wrong formalization.","failure_condition":"The candidate fails if fragment membership is not enforceable, the behavioral predicate cannot faithfully represent the policy, or consequential users cannot act on UNKNOWN without converting it to a Boolean default.","problem_falsifier":"Logs and specifications show only a finite or mechanically bounded artifact class, no universal exact claim, and already-distinct unknown/out-of-scope handling; then there is no computability-boundary problem to solve.","intervention_falsifier":"In the bounded pilot, the router does not reduce forced Boolean defaults or overclaims, produces unusably high UNKNOWN/review rates, or its supposedly sound modes yield verified false clearances under their stated assumptions.","risks":["Formalizing policy may erase contextual meaning or encode unequal treatment.","UNKNOWN may disproportionately burden less-resourced creators.","Attackers may target fragment admission or sandbox boundaries.","False alarms may overwhelm reviewers.","A formally valid record may institutionalize a semantically invalid model."]},"null_rationale":null,"classification":{"candidate_kind":"DOMAIN_TRANSFER","prior_art_status":"UNSEARCHED","evidence_maturity":"HYPOTHESIS"},"revision_change_log":{"revision_kind":"ORIGINAL","prior_problem_id":null,"prior_causal_lever_id":null,"problem_changed":false,"causal_lever_changed":false,"conceptual_changes":[],"operational_changes":[],"repairs_addressed":[]},"confidence":0.78,"generator_notes":"Closed-book mapping. The candidate is conditional on the admitted media class containing programmable behavior and on a faithful nontrivial behavioral-policy predicate; neither undecidability nor practical prevalence is established by the packet."}