AIm³ MentalArena

Best role: turn ambiguous requirements into buildable architecture, implementation gates, and evidence.

1. Legacy-system upgrade: add a major capability or migrate architecture without silently breaking accepted behavior, file formats, APIs, or release workflows.
2. Failed-build diagnosis: explain why code passes smoke tests but fails on real datasets, production traffic, edge cases, or performance limits.
3. Agent/product architecture: decide between a single model, RAG, tools, multi-agent orchestration, or deterministic code, then define the smallest decisive build and eval.

Best role: run blind, cost-aware comparisons and preserve dissent instead of choosing a model by reputation.

1. Model and workflow selection: choose the best model, RAG stack, or agent workflow for coding, support, analysis, or research under quality, latency, cost, and privacy constraints.
2. Orchestration ablation: test whether a multi-agent council actually beats one strong model when the seed, sources, builder, and evaluator are frozen.
3. AI-system risk audit: locate hallucination, prompt-injection, data-leakage, policy, calibration, and human-oversight failures before production promotion.

Best role: keep unconventional ideas alive long enough to convert them into competing mechanisms and kill-tests.

1. New machine or energy concept: split a speculative mechanism into variants, constraints, failure modes, simulations, and the smallest physical prototype.
2. Robotics/control architecture: resolve conflicting sensor, actuator, controller, redundancy, power, safety, and cost choices.
3. Algorithmic invention: explore a large design space, preserve minority hypotheses, and define controls that distinguish a real algorithmic gain from benchmark overfit.

Best role: act as an evidence court across claims, source files, code, logs, and tests.

1. Startup technology claim: verify whether a claimed benchmark, compression gain, AI capability, or trading result survives matched controls and fresh data.
2. Vendor acceptance: compare contract requirements with the delivered repository, test reports, security posture, and missing evidence.
3. Research-package forensics: find leakage, lookahead, impossible accounting, open-tail dependence, cherry-picked runs, or absent negative results in a ZIP bundle.

Best role: generate independent hazard views, merge them into a safety case, and demand testable mitigations.

1. FMEA and hazard analysis: identify coupled failures in a battery system, robot, industrial machine, medical device, or autonomous subsystem.
2. Safety-case construction: connect every safety claim to evidence, assumptions, owner, test, residual risk, and operating boundary.
3. Recovery design: compare graceful degradation, fail-safe shutdown, redundancy, alarm, and human-takeover strategies under cascading faults.

Best role: coordinate competing attack hypotheses and evidence without autonomously executing defensive actions.

1. Incident root cause: decide whether conflicting telemetry supports stolen credentials, application exploit, insider action, supply-chain compromise, or false positive.
2. Threat model: map attack paths through an AI-enabled SaaS product, model supply chain, RAG corpus, tools, identities, and third-party integrations.
3. Remediation priority: rank fixes by exploitability, blast radius, evidence strength, recovery cost, and dependency rather than by the loudest team opinion.

Best role: generate competing hypotheses, explicit falsifiers, and experiments rather than one polished story.

1. Unexplained anomaly: produce mechanistically different explanations, list what each predicts, and design the cheapest experiment that separates them.
2. Replication planning: audit an influential result for hidden degrees of freedom, confounders, missing controls, and a preregistered replication design.
3. Research-program synthesis: turn a fragmented literature into a claim ledger, unresolved conflicts, prioritized hypotheses, and a sequence of empirical tests.

Best role: support research design and evidence review under expert oversight.

1. Mechanism and assay design: compare biological mechanisms, biomarkers, assay choices, off-target risks, and experiments that could reject each route.
2. Contradictory preclinical evidence: reconcile cell, animal, omics, and literature findings while exposing where the datasets genuinely disagree.
3. Trial-protocol review: challenge endpoints, eligibility, stratification, safety monitoring, operational feasibility, and failure-to-learn risks before launch.

Best role: prepare a clinician-led review; never act as an autonomous diagnostician or prescriber.

1. Complex case conference: organize imaging, pathology, labs, history, guidelines, and specialist disagreements for a tumor board or rare-disease meeting.
2. Missing-information analysis: identify which additional test, history item, or specialist opinion would most change the current decision.
3. Guideline conflict: show why recommendations differ across patient subgroups, comorbidities, values, contraindications, and evidence quality.

Best role: build a source and contradiction ledger without inventing missing facts.

1. Conflicting public claims: map who said what, when, on which evidence, and which claims are unsupported or mutually incompatible.
2. Timeline and entity network: connect documents, companies, payments, decisions, people, and events while preserving provenance.
3. Next reporting move: identify the three missing documents, interviews, data requests, or verification steps most likely to resolve the story.

Best role: apply explicit criteria, preserve minority objections, and expose evidence gaps.

1. Patent landscape: compare novelty, claim overlap, obviousness arguments, design-around routes, and uncertain prior art.
2. Grant review: separate scientific importance, methodological quality, feasibility, team capacity, and high-risk/high-reward value.
3. Procurement decision: compare suppliers on total cost, lock-in, interoperability, security, evidence quality, implementation risk, and hidden exclusions.

Best role: turn broad obligations into a traceable evidence matrix and remediation plan.

1. Requirement mapping: map an AI or software system against applicable law, standards, internal policy, risk tier, and responsible owner.
2. Audit-readiness package: connect every control to implementation evidence, test result, exception, residual risk, and review date.
3. Supplier compliance gap: identify what a vendor proves, merely asserts, omits, or cannot contractually guarantee.

Best role: prevent one attractive strategy narrative from crowding out alternatives and kill-tests.

1. Competitive response: model how incumbents, low-cost entrants, regulators, suppliers, and customers could react to a new product.
2. Market-entry decision: compare geographies, segments, pricing, partnerships, operational burden, and downside scenarios.
3. Pre-mortem: identify the smallest early signals that would show a major investment, acquisition, or product thesis is wrong.

Best role: prepare transparent options and tests; democratic legitimacy remains human and institutional.

1. Housing-policy design: compare affordability, supply, tenant protection, fiscal cost, land use, legal constraints, and second-order effects.
2. Transport or health intervention: combine citizen, operator, budget, equity, safety, and implementation perspectives into pilotable options.
3. Policy evaluation: define who should benefit, measurable outcomes, counterfactuals, data needs, failure criteria, and an adaptation path.

Best role: function as a research and risk committee, not an autonomous trading or investment executor.

1. Bull/bear investment case: independently analyze earnings quality, cash flow, balance sheet, competition, valuation, catalysts, and failure modes.
2. M&A due diligence: challenge revenue quality, customer concentration, technology claims, liabilities, integration risk, and downside scenarios.
3. Model/backtest risk: search for leakage, overfit, unrealistic fills, regime dependence, hidden leverage, open-tail exposure, and weak controls.

Best role: support issue spotting and adversarial review; qualified counsel remains responsible.

1. Contract ambiguity: expose conflicting definitions, liability gaps, termination traps, data rights, service levels, and unenforceable assumptions.
2. Litigation preparation: build both sides' strongest arguments, evidence dependencies, credibility risks, and unresolved factual questions.
3. Regulatory interpretation: compare plausible readings, jurisdictional differences, precedent, operational consequences, and questions requiring formal advice.

Best role: integrate engineering, economics, environment, resilience, and public constraints.

1. Grid and storage plan: compare generation, storage, transmission, demand response, reliability, cost, permitting, and extreme-weather performance.
2. Climate-adaptation portfolio: prioritize flood, heat, water, wildfire, coastal, and health measures under uncertain budgets and futures.
3. Infrastructure siting: compare routes or sites on lifecycle cost, safety, ecology, community impact, construction risk, and long-term resilience.

Best role: separate symptoms from root causes and test policies across disruption scenarios.

1. Repeated delivery failure: distinguish forecast error, supplier quality, capacity, scheduling, logistics, incentives, and data problems.
2. Inventory policy: compare service level, working capital, perishability, volatility, lead time, substitution, and shortage cost.
3. Supplier resilience: identify single points of failure, hidden tier dependencies, recovery options, dual-source tradeoffs, and trigger thresholds.

Best role: synthesize conflicting research into testable product choices rather than a feature wish list.

1. Conflicting user evidence: separate real needs, stated preferences, segment differences, accessibility needs, and research artifacts.
2. MVP definition: choose the smallest feature set that tests the core value proposition with explicit success and failure criteria.
3. Edge-case audit: examine onboarding, failure recovery, mobile behavior, accessibility, trust, privacy, and high-stress user journeys.

Best role: combine subject expertise, pedagogy, assessment, accessibility, and teacher judgment.

1. Curriculum redesign: sequence concepts for mixed prior knowledge while preserving depth, practical transfer, and prerequisite structure.
2. Assessment design: measure reasoning and application rather than rote recall or easy AI-generated output.
3. Learning-gap intervention: identify why a learner is stuck and propose targeted exercises, explanations, and teacher checkpoints.

Best role: preserve a distinctive human vision while using independent creative lenses and continuity checks.

1. Story repair: fix a weak second act, predictable twist, flat antagonist, or emotional arc without losing the core theme.
2. Franchise continuity: reconcile canon, character voice, world rules, timeline, visual motifs, and future sequel constraints.
3. Game narrative and mechanics: align player agency, progression, difficulty, economy, story, production cost, and replay value.

Best role: support local experts with scenario breadth and resource tradeoffs; never treat affected communities as mere data.

1. Aid allocation: compare urgency, access, vulnerability, logistics, security, equity, local capacity, and uncertainty.
2. Evacuation and shelter plan: test routes, transport, medical needs, communication failures, weather, congestion, and special populations.
3. Response coordination: reconcile government, NGO, community, medical, logistics, and donor constraints into a staged plan.

Best role: integrate local practice with agronomy, weather, economics, logistics, and environmental constraints.

1. Crop strategy: compare crop mix, soil, water, weather risk, input cost, labor, market demand, and rotation.
2. Pest or disease response: distinguish likely causes, surveillance needs, intervention options, resistance risk, and economic thresholds.
3. Food-chain loss: reduce spoilage and waste across harvest timing, storage, cold chain, transport, processing, and demand planning.

Best role: improve systems and roles; not automate hiring, firing, promotion, or disciplinary decisions.

1. Decision bottleneck: identify unclear authority, duplicated approvals, missing ownership, information delay, and incentive conflict.
2. Role redesign: define responsibilities, interfaces, escalation rules, success measures, and transition risks during growth or reorganization.
3. Organizational postmortem: separate individual blame from process, workload, tooling, communication, and governance failures.

Best role: clarify values, assumptions, options, and experiments; the person retains responsibility.

1. Career change: compare meaning, income, risk, skill gap, location, family effects, reversibility, and a low-cost trial.
2. Relocation or education choice: structure financial, social, legal, health, opportunity, and long-term tradeoffs.
3. Care decision: organize family preferences, medical information, practical support, cost, uncertainty, and questions for professionals.

Better tool: one model, search, or deterministic software. MAL adds avoidable cost and latency.

1. Simple transformation: translate a page, rewrite an email, format a document, or summarize a short supplied text.
2. Deterministic task: calculate a value, validate a schema, rename files, or convert a known data format.
3. Routine classification: route a standard support ticket, answer a stable FAQ, or extract fields from a familiar template.

Better tool: deterministic, tested, certified control systems. MAL may assist design and audit, never the control loop.

1. Immediate physical control: flight control, industrial interlock, robot emergency stop, or power-grid protective relay.
2. Millisecond execution: high-frequency order routing, collision avoidance, or latency-bound actuator control.
3. Emergency clinical signal: life-critical alarm, dosage actuation, ventilator control, or autonomous treatment decision.