Manifold Grace
RESEARCH PROPOSITION · ECONOMICS · AI GOVERNANCE · 5 NOVEMBER 2026

Equitable Positive-Sum Growth: Mapping Where It Can, Cannot and Might Work

A prospective framework for testing the feasibility of shared economic benefit, disclosing uncertainty and identifying unacceptable harm—not presuming that a win–win solution exists in every case.

Core proposition and research perimeter

Investigate whether and where economically productive arrangements can increase value without imposing avoidable, unmitigated losses on others; make the feasibility conditions, affected populations and evidence boundaries publicly inspectable. Where such arrangements are demonstrated, investigate the safeguards that could make them routine and resilient. Where they cannot be demonstrated, disclose the limits, who may lose, and what remedies or alternatives remain available.

This proposal is a conceptual research specification attributed to Manifold Grace; it is not an empirically established universal theory, existing regulatory standard, AI-validated global map or proof that every conflict can be rendered positive-sum. Research should examine both Pareto improvements (no included participant worse off), distributional or equity objectives (fairness of outcomes across groups), and minimum living-standard constraints. Equal absolute benefits, equal proportional improvements and priority to those with greatest needs are different normative aims: they must be selected transparently rather than smuggled into a single word, “equal”.

Cases and boundaries: a “case” is a defined policy, product or allocation design within a stated jurisdiction, time horizon, population, alternative baseline, evidence set and constraints. New versions or changed assumptions create a new assessment, not a permanent verdict on all uses of a technology.

Three evidence-status and risk zones

Green · demonstrated within scope

An option satisfies defined welfare, legal, distributional and risk thresholds on adequate evidence, including relevant groups and side effects. No material, unmitigated harms identified within the declared assessment scope.

Decision: eligible for controlled use with monitoring, periodic review, remedy and rollback. Green is never a permanent all-clear or global certification.

Amber · uncertain or conditional

Benefits are plausible but evidence is incomplete, assumptions are sensitive, stakeholder impacts conflict, or necessary safeguards are untested.

Decision: restrict to proportionate research, independent review or controlled pilots where lawful and safe; disclose unanswered questions, avoid promoting as validated win–win.

Red · unacceptable harm or violated constraints

Credible evidence establishes material unmitigated loss, unlawful conduct, violations of essential rights or services, or risk beyond the selected acceptable boundary. An infeasible solution may also be marked red for a specific proposed design, with its reason stated.

Decision: do not deploy or continue the unsafe design under those conditions; disclose affected parties, halt/modify as appropriate, provide remedy and assess alternatives.

Critical distinction: missing evidence normally means amber, not green. Credible serious danger can warrant a provisional red precaution while facts are investigated. Red denotes danger of an option, not the character or intrinsic worth of a person, occupation or community. An analytically impossible win–win within fixed constraints is not automatically misconduct; it must be explained as a feasibility limit and addressed through fair choice, compensation or redesign.

Case-by-case AI-assisted feasibility and ethical loss mapping

StageRequired inquiryDocumentable output
0 · Define values and scopeSpecify affected parties, minimum living standards, fairness criterion, time, region, baseline and rights constraints.Versioned case charter with agreed inclusion/exclusion boundaries.
1 · Historical prior artIdentify previous models, failures, legislation, policy studies, empirical counterexamples and competing theories.Primary-source literature and alternative-design register.
2 · Distribution and exposureMap beneficiaries, payers, workers, bystanders and future impacts, including hidden environmental and unpaid-labour costs.Group-level gain/loss and risk distribution matrix.
3 · Feasible setTest whether all designated constraints can be satisfied simultaneously; distinguish logical impossibility from missing data or insufficient technology.Optimisation or modelling specification with documented assumptions and infeasibility certificates where applicable.
4 · Alternative architecturesCompare cooperation, competition, regulation, public provision, negotiated compensation and mixed systems.Traceable alternative designs and counterfactual comparisons.
5 · AI and human evaluationUse appropriate forecasting, optimisation, causal inference and simulation; independently inspect data, objectives, subgroup effects and model errors.Calibration, uncertainty intervals, external validation and human challenge record.
6 · Green/amber/red decisionApply predeclared benefit, harm, equity, legal and safety thresholds. Consider severe tail risk and reversibility, not merely average net gain.Evidence-bounded zone with reasons, exceptions and named accountability.
7 · Governance and remediationFor green, monitor and review; for amber, narrow exposure and gather evidence; for red, halt, redesign, compensate or remedy as required.Control register, escalation route and corrective-action history.
8 · Long-horizon recheckLook for drift, second-order impacts, transfer of losses to other regions and changed laws or resources.As-of-date reassessment; promotion/demotion decisions with evidence.
9 · Publish and transferRelease methods, source licences, uncertainty, adverse findings and what successor researchers can repeat.Versioned case report and research-transfer dossier.

Illustrative sectors: conditions to investigate, not observed results

The following examples do not label real locations, businesses or people green, amber or red. They show what a genuine case study must investigate before any zone may be published.

1. Food distribution Illustrative · unassessed
Candidate pathway
Reduce spoilage through improved demand forecasts and voluntary producer/buyer coordination
Possible green evidence
Potential food savings, lower waste and more reliable stocks
Amber uncertainty
Amber until verified: farmers' bargaining power, worker hours, price pass-through and nutrition access
Red trigger
Red if coercive purchasing, predatory pricing or foreseeable loss of essential food
Who must be considered
Retailers, growers, workers, low-income households, charities

Status: This is an example of screening criteria, not an actual empirical zone assignment.

2. AI-supported workforce productivity Illustrative · unassessed
Candidate pathway
Use task automation to release time for higher-value work and negotiated skills progression
Possible green evidence
Possible employer productivity and worker benefit
Amber uncertainty
Amber: wage-sharing and employment effects unknown; need longitudinal studies
Red trigger
Red if profits rise while workers lose necessary income without adequate protection
Who must be considered
Employees, contractors, business owners, customers

Status: This is an example of screening criteria, not an actual empirical zone assignment.

3. Local energy flexibility Illustrative · unassessed
Candidate pathway
Optimise demand-response against real technical and household constraints
Possible green evidence
Possible lower peak costs, reduced emissions and grid resilience
Amber uncertainty
Amber: meter data privacy and burden distribution unverified
Red trigger
Red if medically or financially vulnerable households lose essential service
Who must be considered
Households, utilities, vulnerable residents, regulators

Status: This is an example of screening criteria, not an actual empirical zone assignment.

4. Inclusive credit assessment Illustrative · unassessed
Candidate pathway
Evaluate additional lawful information with human appeal and independent fairness review
Possible green evidence
Potential access for qualified excluded borrowers
Amber uncertainty
Amber: proxy discrimination and debt affordability remain uncertain
Red trigger
Red if exploitative interest, privacy violation or unsustainable debt risk
Who must be considered
Applicants, lenders, excluded communities, supervisory authorities

Status: This is an example of screening criteria, not an actual empirical zone assignment.

5. Public-service scheduling Illustrative · unassessed
Candidate pathway
Optimise queues while maintaining service eligibility, accessibility and equal access
Possible green evidence
Possible waiting-time reductions and more effective staff allocation
Amber uncertainty
Amber: unmeasured subgroup waiting times or staff workload
Red trigger
Red if urgent or protected-needs groups are systematically deprived of essential provision
Who must be considered
Service users, frontline staff, commissioners

Status: This is an example of screening criteria, not an actual empirical zone assignment.

6. Research-to-enterprise contributor remuneration Illustrative · unassessed
Candidate pathway
Fund open research through opted-in contracts and transparent paid services
Possible green evidence
Possible paid work and reinvestment without closing public learning
Amber uncertainty
Amber: revenue, demand, ownership and employment classification unknown
Red trigger
Red if worker status is mislabelled or pay unlawfully withheld
Who must be considered
Researchers, contributors, clients, public beneficiaries

Status: This is an example of screening criteria, not an actual empirical zone assignment.

Minimum ethical disclosure when loss is foreseeable

Every assessment should identify who might lose, what might be lost, whether the loss is temporary or irreversible, whether it is avoidable, whether the affected party consents or has a choice, whether compensation is real and accessible, and who bears responsibility for remedy. Publication must not disclose confidential information, identifiable vulnerable individuals or sensitive operational detail without a lawful basis. A public summary may be accompanied by restricted evidence and independently auditable confidential records.

It is insufficient to report “net gain” where loss of housing, food, essential care, lawful work or basic rights is hidden by large gains elsewhere. Conversely, evidence of a trade-off must be documented rather than taken as proof that another party intended harm. The output should distinguish observed adverse effect, credible forecast risk, scenario assumption and unresolved allegation.

Research integrity: an AI-generated zone is a decision-support suggestion until the underlying case is validated. A predictive model cannot ethically guarantee future outcomes, and a simulation cannot establish real-world causation on its own. Human decisions remain accountable and contestable.

Making equitable practice the norm without freezing mistakes

“Gridlock as the norm” is interpreted here as institutionalising verified safeguards through procurement criteria, service-level requirements, impact assessments, contracts, audit trails, monitoring, transparency and correction rights—not forbidding reconsideration or allowing algorithmic classifications to become unchallengeable. Green assessments must expire or be reviewed when evidence, technology, parties, environmental conditions or legal standards change. For amber and red, preserve unsuccessful or infeasible alternatives so later researchers can understand both the obstacle and potential ways around it.

Government appraisal guidance already considers distributional effects and uncertainty; the proposed contribution is an integrated, case-addressable, AI-assisted feasibility-and-loss disclosure map across stages, not an assertion that the underlying ideas of Pareto efficiency, distributional appraisal or traffic-light risk triage were invented here.

Existing foundations and Manifold Grace research lineage

Contribution and status: Manifold Grace conceptual proposal, 5 November 2026, drafted with AI assistance. The mapping method and sample zone triggers are prospective, unvalidated proposals. No empirical green-zone certification, global case census, quantified win–win guarantee, organisational endorsement or completed economic experiment is asserted. Prior economic and governance concepts remain attributable to their established literatures.