{"schema_version":1,"assessment_id":"eoa_inverse_innovation_exp03_opportunity320_20260801","source_experiment_id":"eoa_inverse_innovation_exp03_full320_20260801","cell_id":"computability_boundary_mapping__psychology","archetype_slug":"computability_boundary_mapping","domain_slug":"psychology","title":"Model-relative reachability boundaries for executable psychological models","opportunity_summary":"Evaluate whether an explicitly declared executable psychological-model class supports exact terminating reachability analysis, enforce any proven decidable fragment, and preserve labeled UNKNOWN outcomes for cases handled only by abstraction or bounded simulation. The opportunity is conditional on finding projects that actually make universal Boolean guarantees; the packet establishes neither their prevalence nor a theorem for a concrete model language.","adopter_authorizer":"Executable-model or digital-intervention owners could adopt the research workflow with approval from an independent formal-methods reviewer. Any clinical or human-facing use would additionally require the relevant clinical, ethics, and institutional authorities.","scores":{"meaningful_impact":{"score":3,"rationale":"False clearance of harmful modeled trajectories or rejection of useful models could matter substantially, but the packet provides no evidence about affected-project prevalence, frequency of collapsed statuses, or realized downstream harm."},"stakeholder_pull":{"score":2,"rationale":"The candidate identifies model developers, reviewers, clinicians, and represented people as stakeholders, but supplies no interviews, requests, incidents, procurement signal, or other evidence that an adopter currently prioritizes this problem."},"incremental_advantage":{"score":3,"rationale":"Contract matching, independently checked certificates, enforced fragments, and explicit UNKNOWN directly improve on timeout-as-NO and indiscriminate simulation, but no comparative result shows fewer false Boolean conclusions or sufficient retained coverage."},"distinctiveness_plausibility":{"score":2,"rationale":"The composition is coherent and adapted to executable psychological models, but prior-art status is explicitly UNSEARCHED and novelty relative to formal verification and model-governance practices is unsupported."},"technical_implementability":{"score":3,"rationale":"An offline synthetic pilot on one frozen language and property is bounded and technically conceivable, but no language, semantics, reduction, constructive analyzer, abstraction, or enforcement mechanism has yet been instantiated."},"adoption_authority_feasibility":{"score":4,"rationale":"The proposal identifies model owners and an independent reviewer for research guarantees and separately reserves clinical decisions to clinical, ethics, and institutional authority; feasibility remains conditional on securing those participants."},"evidence_readiness":{"score":2,"rationale":"Evidence maturity is HYPOTHESIS: the packet contains a procedure and falsifiers but no concrete proof artifact, benchmark results, deployed-language audit, downstream-interface test, or demand evidence."},"safety_net_benefit":{"score":4,"rationale":"Preserving UNKNOWN, separating model claims from claims about people, prohibiting patient-care changes, and withdrawing unsupported guarantees could reduce false certainty; benefit depends on labels surviving downstream interfaces and abstractions remaining sound."},"scalability":{"score":3,"rationale":"The contract-and-routing pattern could transfer across projects, but each model language, semantics, target property, reduction, decidable fragment, and abstraction may require substantial bespoke analysis."}},"score_confidence":"LOW","costs":{"first_evidence":{"band_2026_usd":"50K_TO_250K","scope":"One offline pilot covering language and property specification, expressiveness audit, attempted restricted-fragment analyzer, attempted checked reduction, synthetic comparison of exact, abstract, and bounded outputs, and independent review.","confidence":"LOW","assumptions":["One existing model language and toolchain can be selected without major licensing or data expense.","The pilot uses synthetic models and no patient data or live clinical workflow.","The work requires both formal-methods and computational-psychology expertise.","Proof attempts may end unresolved without implying a boundary verdict."]},"initial_deployment_startup":{"band_2026_usd":"250K_TO_1M","scope":"Prepare one research organization to enforce the proven fragment boundary, integrate labeled fallbacks and provenance, build regression tests, and establish independent review and governance procedures.","confidence":"LOW","assumptions":["A pilot has produced usable proof or analysis artifacts.","Deployment is limited to one declared model language and one organizational workflow.","Existing execution and review infrastructure can be adapted rather than replaced.","Clinical decision support remains outside scope."]},"operational_launch":{"band_2026_usd":"250K_TO_1M","scope":"Harden and launch the boundary-routing workflow for routine research use, including interface safeguards for UNKNOWN, documentation, training, security and compliance review, audit trails, and monitored rollback.","confidence":"LOW","assumptions":["No patient-care change or person-level safety classification is authorized.","The organization can enforce language restrictions and prevent bypass through extensions or external services.","Launch includes evaluation of false Boolean conclusions and useful coverage.","Unexpected semantic or integration work could move the effort outside this band."]},"annual_recurring":{"band_2026_usd":"50K_TO_250K","scope":"Maintain proof and analyzer artifacts, review language or property changes, run regression and interface audits, monitor UNKNOWN handling, retrain users, and retain independent review capacity for one deployment.","confidence":"LOW","assumptions":["The model language and target property change infrequently.","The deployment remains within one organization and a small number of workflows.","Material language extensions or new properties would be treated as new projects rather than routine maintenance."]}},"research_burden":"HIGH","earliest_credible_horizon":"3_TO_12_MONTHS","pipeline_gates":{"recognizable_externally_supportable_problem":{"status":"UNCERTAIN","reason":"The packet describes a coherent failure mode and falsifier but contains no observed project, incident, prevalence evidence, or adopter testimony showing that universal exact terminating claims are actually being made."},"identifiable_adopter_or_authorizer":{"status":"YES","reason":"Model owners and an independent formal-methods reviewer are identified for research guarantees, with clinical, ethics, and institutional authorities explicitly retained for human-facing use."},"distinct_testable_incremental_claim":{"status":"YES","reason":"The candidate makes a separable claim that enforced boundary routing and explicit UNKNOWN will reduce false Boolean conclusions relative to timeout or bounded-simulation baselines while retaining useful coverage."},"bounded_next_evidence_step":{"status":"YES","reason":"The sealed candidate authorizes an offline pilot on one declared language and reachability property, with synthetic comparisons, independent proof checking, and explicit falsifiers."},"no_unresolved_safety_or_authority_stop":{"status":"YES","reason":"The first step is offline; patient-care changes and person-level danger or safety labels are excluded; and semantic loss, failed review, unenforceable boundaries, or UNKNOWN-as-NO trigger withdrawal."},"implementation_cost_scope_and_range":{"status":"UNCERTAIN","reason":"The pilot scope is bounded enough for a broad resource band, but no concrete language, codebase, organization, proof complexity, integration surface, compliance context, or deployment scale supports a reliable implementation range."}},"blocking_evidence":["No concrete executable psychological-model language, frozen semantics, or reachability property has been selected.","No independently checked total-analyzer certificate or contract-matched impossibility reduction is provided.","No evidence establishes that target projects make universal exact terminating claims or collapse timeout and UNKNOWN into Boolean verdicts.","False Boolean conclusions, useful coverage, acceptable UNKNOWN rates, and downstream label preservation lack predefined operational measures.","Novelty and overlap with existing formal-verification and governance approaches are unmeasured."],"next_evidence_step":"Run the authorized offline pilot on one frozen executable model language and one formalized reachability property: independently check both the restricted-fragment analyzer attempt and the unrestricted-class reduction attempt, then compare boundary routing against the current bounded or timeout-based method on preregistered synthetic cases labeled YES, NO, UNKNOWN, out-of-scope, nontermination, and model failure. Falsify progression if no universal-claim problem is observed, the formal contracts cannot be matched, or routing does not reduce false Boolean outputs while meeting a predefined useful-coverage threshold.","research_questions":["Does a specific prospective adopter make an exact, class-wide, always-terminating reachability claim, and how are timeout, failure, and out-of-scope cases currently reported?","What are the frozen syntax, semantics, quantifiers, model class, and reachability property for the selected language?","Can an independent reviewer validate either a total correct analyzer for a useful fragment or a contract-matched impossibility reduction for the broader class?","Compared with the current method, does boundary routing reduce false Boolean conclusions without an unusable UNKNOWN rate or loss of practically essential model cases?","Do downstream interfaces preserve UNKNOWN, scope, and guarantee labels rather than converting them into person-level or clinical certainty?","What existing methods or governance practices already provide the same contract specification, proof checking, fragment enforcement, and labeled fallback behavior?"],"recommendation":"PARTNERED_RESEARCH","uncertainty_constraints":["Closed-book assessment: no external validation of prevalence, demand, novelty, market size, prior art, realized impact, or exact cost was available.","All benefit claims apply to formal executable model-property pairs, not to the computability, predictability, safety, or behavior of real people.","Undecidability remains unestablished until a concrete language and property support an independently checked, contract-matched reduction.","Decidability of many finite, bounded, or statistical psychological models could falsify the targeted problem for relevant adopters.","Cost bands are resource-equivalent planning ranges conditioned on a single-language, offline-to-research deployment and could change materially with proof complexity or clinical integration."],"closed_book_prior_art_boundary":"Prior-art status is UNSEARCHED. This assessment makes no claim that the proposed composition is novel, rare, or absent from formal verification, computational modeling, digital-intervention governance, or adjacent practice; external prior-art research is required before any originality claim."}