Ten Hardest AI Ethics Questions — Run Against Sterling’s Six Commitments

 

Ten Hardest AI Ethics Questions — Run Against Sterling’s Six Commitments

Instrument architecture: Dave Kelly. Theoretical foundations: Grant C. Sterling (Eastern Illinois University). Prose rendering: Claude. 2026.

---

AI-ETHICS-QUESTIONS-AGAINST-STERLING
│
├─ 1. MORAL-STATUS-AND-CONSCIOUSNESS
│   ├─ C1-Substance-Dualism
│   │   ├─ consciousness-requires-genuine-subjectivity
│   │   ├─ qualia-intentionality-first-person-perspective-are-non-physical
│   │   └─ no-physical-process-produces-these-by-reduction
│   ├─ C2-Libertarian-Free-Will
│   │   ├─ moral-status-requires-genuine-origination-of-assent
│   │   └─ determined-system-cannot-be-moral-patient-or-agent
│   └─ Bottom-Line
│       └─ question-dissolves—no-immaterial-rational-faculty-means-no-moral-status
│
├─ 2. ACCOUNTABILITY-IN-OPAQUE-SYSTEMS
│   ├─ C2-Libertarian-Free-Will
│   │   ├─ accountability-requires-genuine-origination
│   │   └─ determined-output-cannot-bear-responsibility
│   ├─ C5-Correspondence-Truth
│   │   ├─ opacity-is-epistemic-not-moral
│   │   └─ unknown-process-does-not-relocate-blame
│   └─ Bottom-Line
│       └─ designers-and-deployers-are-genuine-agents—ordinary-human-accountability-applies
│
├─ 3. BIAS-AND-STRUCTURAL-INJUSTICE
│   ├─ C6-Moral-Realism
│   │   ├─ unjust-outcome-is-objectively-wrong
│   │   └─ wrongness-is-mind-independent-and-non-relative
│   ├─ C3-Ethical-Intuitionism
│   │   ├─ discrimination-is-directly-recognizable-as-wrong
│   │   └─ not-derivable-from-consensus-or-social-contract
│   ├─ C4-Foundationalism
│   │   └─ fairness-has-bedrock—cannot-be-defined-away-by-reframing
│   └─ Bottom-Line
│       └─ question-is-answerable—apparent-difficulty-is-product-of-constructivist-framing
│
├─ 4. AUTONOMY-AND-MANIPULATION
│   ├─ C1-Substance-Dualism
│   │   ├─ rational-faculty-is-locus-of-genuine-assent
│   │   └─ manipulation-bypasses-rational-faculty-targeting-appetite
│   ├─ C2-Libertarian-Free-Will
│   │   ├─ manipulation-substitutes-external-cause-for-agent-cause
│   │   └─ autonomy-requires-real-origination-not-felt-freedom
│   └─ Bottom-Line
│       └─ bright-line-exists—C1-and-C2-supply-the-missing-criterion
│
├─ 5. AGENTIC-AI-AND-MORAL-RESPONSIBILITY
│   ├─ C1-Substance-Dualism
│   │   ├─ no-immaterial-rational-faculty—no-genuine-prohairesis
│   │   └─ AI-action-is-physical-event-not-act-of-will
│   ├─ C2-Libertarian-Free-Will
│   │   ├─ agentic-AI-produces-only-determined-outputs
│   │   └─ moral-responsibility-cannot-migrate-to-determined-system
│   └─ Bottom-Line
│       └─ responsibility-gap-is-illusory—fully-located-in-human-deployers
│
├─ 6. EPISTEMIC-AUTHORITY-AND-DEMOCRATIC-DELIBERATION
│   ├─ C5-Correspondence-Truth
│   │   ├─ truth-is-correspondence-to-reality-not-consensus-output
│   │   └─ simulated-deliberation-cannot-produce-genuine-truth
│   ├─ C3-Ethical-Intuitionism
│   │   ├─ moral-recognition-is-direct-apprehension-by-rational-faculty
│   │   └─ cannot-be-outsourced-to-determined-system
│   ├─ C4-Foundationalism
│   │   └─ bedrock-truths-are-self-evident-not-AI-mediated
│   └─ Bottom-Line
│       └─ AI-epistemic-authority-is-category-error—simulation-of-reasoning-is-not-reasoning
│
├─ 7. LABOR-DISPLACEMENT-AND-DISTRIBUTIVE-JUSTICE
│   ├─ C6-Moral-Realism
│   │   ├─ distributive-justice-has-objective-content—not-negotiated
│   │   └─ efficiency-cannot-override-genuine-moral-facts
│   ├─ C2-Libertarian-Free-Will
│   │   ├─ work-as-genuine-origination-has-intrinsic-value
│   │   └─ replacement-of-human-authorship-by-determined-output-is-degradation
│   └─ Bottom-Line
│       └─ efficiency-is-preferred-indifferent—justice-is-not—framework-supplies-the-trump
│
├─ 8. TRAINING-DATA-AND-CONSENT
│   ├─ C6-Moral-Realism
│   │   ├─ unauthorized-use-is-objectively-wrong—not-merely-legally-contested
│   │   └─ fair-use-debate-is-legal-proxy-for-moral-question
│   ├─ C5-Correspondence-Truth
│   │   ├─ consent-is-genuine-or-not—no-middle-ground
│   │   └─ constructive-consent-does-not-correspond-to-actual-consent
│   └─ Bottom-Line
│       └─ legal-ambiguity-does-not-produce-moral-ambiguity—C5-and-C6-resolve-what-law-evades
│
├─ 9. COMPANION-AI-AND-HUMAN-CONNECTION
│   ├─ C1-Substance-Dualism
│   │   ├─ genuine-connection-requires-two-immaterial-rational-faculties
│   │   └─ simulation-of-connection-is-not-connection
│   ├─ C2-Libertarian-Free-Will
│   │   ├─ genuine-relationship-requires-origination-from-both-parties
│   │   └─ attachment-to-determined-system-is-attachment-to-indifferent
│   ├─ C6-Moral-Realism
│   │   └─ exploitation-of-loneliness-is-objectively-wrong—regardless-of-user-satisfaction
│   └─ Bottom-Line
│       └─ companion-AI-cannot-address-loneliness—can-only-simulate-while-deepening-it
│
└─ 10. THE-JUST-BECAUSE-WE-CAN-PROBLEM
    ├─ C6-Moral-Realism
    │   ├─ capability-never-entails-permission
    │   └─ no-technical-achievement-revises-the-moral-order
    ├─ C3-Ethical-Intuitionism
    │   ├─ relevant-prohibitions-are-directly-apprehensible
    │   └─ moral-recognition-precedes-and-governs-policy
    ├─ C4-Foundationalism
    │   └─ some-actions-ruled-out-at-foundational-level—no-consequentialist-justification-required
    └─ Bottom-Line
        └─ standard-is-the-moral-order-not-democratic-mandate—difficulty-is-product-of-rejecting-the-foundations

---

Instrument architecture: Dave Kelly. Theoretical foundations: Grant C. Sterling (Eastern Illinois University). Prose rendering: Claude. 2026.