{"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__biology_ecology","archetype_slug":"computability_boundary_mapping","domain_slug":"biology_ecology","title":"Model-Relative Solvability Boundaries for Long-Run Ecosystem Forecasts","opportunity_summary":"Evaluate whether formal model classification, restricted exact solvers, sound abstractions, and explicit UNKNOWN routing can prevent unsupported long-run extinction or persistence labels. The synthetic-model test is coherent and safely bounded, but practical demand, ecological formalization, proof obligations, workflow effects, and distinctiveness remain unverified.","adopter_authorizer":"The accountable conservation program owner may authorize synthetic analysis; an independent proof approver must authorize formal guarantees, while existing ecological authorities and affected-party processes retain authority over real interventions.","scores":{"meaningful_impact":{"score":4,"rationale":"If conservation decisions currently receive authoritative labels that conflate timeout, failure, and unresolved cases, guarantee-aware routing could prevent consequential false certainty and wasted solver effort. The packet does not establish how often this problem occurs."},"stakeholder_pull":{"score":2,"rationale":"Potential users and affected parties are identified, but no adopter request, observed workflow demand, or prevalence evidence is supplied. The candidate itself notes that many practical questions may instead be finite-horizon or dominated by uncertainty and cost."},"incremental_advantage":{"score":3,"rationale":"The proposed fragment checker, checked guarantees, and explicit UNKNOWN state offer a testable advantage over interpreting timed simulations as long-run answers. Advantage remains uncertain because no instantiated solver, ecological language, benchmark result, or downstream-use evidence is included."},"distinctiveness_plausibility":{"score":2,"rationale":"The mechanism stack is specific, but prior art is explicitly unsearched and related verification, reachability, population-process, and formal decision-support work has not been compared. Distinctiveness therefore lacks affirmative evidence."},"technical_implementability":{"score":3,"rationale":"A synthetic finite-state comparison against exhaustive ground truth is technically bounded, and halt conditions are defined. However, executable ecological syntax and semantics, useful outcome predicates, mechanical fragment recognition, solver artifacts, and any property-preserving reduction remain incomplete."},"adoption_authority_feasibility":{"score":4,"rationale":"The packet separates program ownership, independent proof approval, ecological intervention authority, and affected-party processes. Feasibility is favorable for a synthetic study, although adoption into operational decision workflows has not been demonstrated."},"evidence_readiness":{"score":3,"rationale":"The candidate supplies a comparison setting, falsifiers, exclusions, and rollback rules, enabling study design. Readiness is limited because the pilot language, benchmark corpus, restricted solver, abstraction, and proof obligations are not yet instantiated."},"safety_net_benefit":{"score":5,"rationale":"Explicit UNKNOWN and out-of-scope states, prohibition of field action, independent proof review, conflict checks against exhaustive ground truth, and guarantee withdrawal directly limit overclaiming and provide a strong rollback path."},"scalability":{"score":2,"rationale":"Guarantees must be versioned and tied to enforceable model fragments, while each richer language or ecological predicate may require new formalization and proof. A useful fragment could also exclude the interactions stakeholders need."}},"score_confidence":"MODERATE","costs":{"first_evidence":{"band_2026_usd":"50K_TO_250K","scope":"Formalize one pilot ecosystem-model language and outcome predicate, implement a restricted solver and explicit-UNKNOWN router, construct a synthetic finite-state benchmark, compare results with exhaustive ground truth and a timed-simulation baseline, and obtain independent proof review.","confidence":"MODERATE","assumptions":["The study remains entirely synthetic and requires no field data or intervention.","A compact finite-state language and benchmark can be implemented with a small interdisciplinary formal-methods and ecology team.","The unrestricted impossibility result is not required for the constructive pilot.","Existing general-purpose development and verification infrastructure is usable."]},"initial_deployment_startup":{"band_2026_usd":"250K_TO_1M","scope":"Harden the checker, solver, abstraction, provenance records, guarantee versioning, and interface controls for a single conservation-program analysis workflow, including validation, governance design, training, and independent review.","confidence":"LOW","assumptions":["Deployment covers one organization and a narrow enforceable model fragment.","No field-management decision is automated.","Integration complexity and compliance requirements are not specified in the packet.","Ecological model translation requires substantial specialist coordination."]},"operational_launch":{"band_2026_usd":"250K_TO_1M","scope":"Launch a governed analytical service for a bounded model class, with access controls, audit trails, user training, downstream preservation of UNKNOWN states, incident handling, and approval procedures for each guarantee version.","confidence":"LOW","assumptions":["Launch is limited to decision support rather than autonomous conservation action.","One bounded model family has passed synthetic and workflow validation.","Independent proof approval and affected-party governance can use existing institutional processes.","No major new computing equipment or proprietary data acquisition is required."]},"annual_recurring":{"band_2026_usd":"50K_TO_250K","scope":"Maintain software and benchmarks, review model-language or assumption changes, renew guarantee approvals, monitor suppression or misuse of UNKNOWN outputs, train users, and withdraw invalidated guarantees.","confidence":"LOW","assumptions":["Operations remain limited to one narrow model family and a small number of workflows.","Material language changes trigger additional project work outside routine maintenance.","Monitoring can rely primarily on retained audit records rather than field outcome attribution.","No unresolved reduction is represented as an operational guarantee."]}},"research_burden":"HIGH","earliest_credible_horizon":"3_TO_12_MONTHS","pipeline_gates":{"recognizable_externally_supportable_problem":{"status":"UNCERTAIN","reason":"The packet clearly specifies the possible collapse of timeout, failure, out-of-scope, and unresolved states into authoritative labels, but explicitly treats the existence and practical prevalence of a universal-prediction requirement as a hypothesis requiring external evidence."},"identifiable_adopter_or_authorizer":{"status":"YES","reason":"The accountable conservation program owner is identified as the analysis authorizer, independent proof approval is required for formal claims, and real interventions remain under existing ecological and affected-party authority."},"distinct_testable_incremental_claim":{"status":"YES","reason":"The proposal can test whether a mechanically enforced restricted solver and explicit-UNKNOWN routing avoid unsafe long-run labels relative to exhaustive ground truth and the timed-simulation baseline."},"bounded_next_evidence_step":{"status":"YES","reason":"The candidate authorizes a synthetic finite-state study only, names exhaustive ground truth as the comparison, and supplies concrete falsifiers and halt conditions."},"no_unresolved_safety_or_authority_stop":{"status":"YES","reason":"No field action is needed for the first study; proof and intervention authority are separated, affected parties are named, unsafe exact labels are prohibited, and guarantees can be withdrawn on failure."},"implementation_cost_scope_and_range":{"status":"YES","reason":"The synthetic evidence step and a narrow single-program deployment can be scoped from the specified solver, checker, routing, review, governance, and rollback components, although the resulting broad resource bands have limited confidence."}},"blocking_evidence":["Evidence that an identifiable ecology or conservation workflow actually requests or behaves as though it has a universally correct, always-terminating long-run predictor.","An executable pilot language with scientifically useful semantics, formal outcome predicates, and mechanically decidable fragment membership.","Benchmark evidence that every safe verdict from the restricted solver or abstraction agrees with exhaustive finite-state ground truth and that excluded cases route to UNKNOWN.","A checked property-preserving reduction before making any unrestricted impossibility claim; otherwise that claim must remain unresolved.","Workflow evidence that downstream interfaces and users preserve UNKNOWN and do not convert it into an authoritative ecological label.","A prior-art comparison sufficient to delimit any distinctiveness claim."],"next_evidence_step":"On a fixed, preregistered corpus of synthetic finite-state ecosystem models, compare the restricted solver and sound abstraction with exhaustive ground truth and a timed-simulation baseline. Proceed only if the fragment checker rejects every excluded model, all issued safe verdicts match ground truth, and UNKNOWN remains explicit downstream; any conflicting verdict, leaked excluded model, or open proof obligation falsifies readiness and triggers guarantee withdrawal.","research_questions":["Do conservation programs actually seek universal long-run guarantees, or are their requirements bounded to finite horizons, finite model classes, or explicitly heuristic forecasts?","Can a scientifically useful ecological model language and extinction, persistence, or stability predicates be formalized while retaining mechanically recognizable solvable fragments?","Does the restricted solver remain sound and useful against exhaustive ground truth, and how often does it return UNKNOWN or coarse false alarms?","Can any proposed impossibility encoding satisfy totality, computability, and ecological-predicate preservation under independent proof review?","Will operational interfaces and governance preserve UNKNOWN, model-relative guarantees, and affected-party authority when decisions are made?","How does the candidate differ from existing ecological-model verification, dynamical reachability, population-process analysis, and formal decision-support approaches?"],"recommendation":"PARTNERED_RESEARCH","uncertainty_constraints":["Closed-book evaluation provides no evidence of problem prevalence, stakeholder demand, market size, realized impact, prior art, or world novelty.","The ecological language, semantics, outcome predicates, fragment checker, restricted solver, abstraction, and benchmark are proposed rather than completed.","No undecidability or other unrestricted impossibility result has been established; analogy and timeout cannot substitute for a checked reduction.","Formal truth about a model does not establish truth about an open, stochastic, or misspecified ecosystem.","Cost bands are resource-equivalent planning ranges based on the described scope, not observed quotations or deployment histories.","Scalability and adoption depend on whether useful stakeholder cases remain inside an enforceable fragment and whether downstream systems retain UNKNOWN."],"closed_book_prior_art_boundary":"Prior art is explicitly unsearched. This assessment makes no claim that the approach, its components, or their combination is novel, uncommon, or absent from ecological verification, reachability analysis, population-process methods, or formal decision-support practice."}