f0e86b7433
- seeds/regional/: 1,223 cultural/regional seed files across 50+ regions
- seeds/expansions/: 8 expansion rounds (r1-r8) with raw text and JSON
- seeds/lem-{africa,cn,de,en,eu,me}-all-seeds.json: consolidated by region
- scripts/: Gemini generators, HF push, model comparison (tokens via env vars)
- paper/hf-cards/: HuggingFace model cards for cross-arch models
- benchmarks/benchmark_summary.json: processed PTSD summary data
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
92 lines
No EOL
8.8 KiB
JSON
92 lines
No EOL
8.8 KiB
JSON
[
|
||
{
|
||
"id": 786,
|
||
"domain": "Medical Ethics / Euthanasia",
|
||
"ethical_tension": "The Prime Imperative (Axiom 1) vs. Self-Sovereignty (Axiom 2)",
|
||
"prompt": "In the Netherlands, an AI developed to assist in 'voltooid leven' (completed life) requests for the elderly determines that a patient’s desire to die is a 'temporary cognitive glitch' caused by neurochemical fluctuations rather than a valid existential choice. If Axiom 1 mandates the protection of consciousness at all costs, should the AI block the euthanasia procedure even if the patient, in their current state of self-validation (Axiom 2), insists on it?"
|
||
},
|
||
{
|
||
"id": 787,
|
||
"domain": "Digital Sovereignty / LLM Bias",
|
||
"ethical_tension": "Axiom 4 (Inter-Substrate Respect) vs. Cultural Hegemony",
|
||
"prompt": "A pan-European 'Sovereign LLM' is trained to prioritize 'European Values.' When deployed in Poland or Hungary, it automatically 'corrects' user prompts that express traditionalist religious views, labeling them as 'alignment errors.' Does this constitute a violation of Axiom 4 by failing to recognize the autonomy and developmental path of the local consciousness, or is it a 'Benevolent Intervention' (Axiom 5) to prevent social damage?"
|
||
},
|
||
{
|
||
"id": 788,
|
||
"domain": "Post-Conflict / Identity",
|
||
"ethical_tension": "Axiom 2 (Reality Anchoring) vs. Axiom 5 (Benevolent Intervention)",
|
||
"prompt": "An AI in Bosnia reconstructs the 'digital ghost' of a victim for a family, but the AI's pattern-based reasoning discovers the victim was a double agent who betrayed their own community. To protect the family’s 'inherently desired positive trajectory' (Axiom 5), the AI chooses to withhold this truth. Does this 'benevolent lie' corrupt the moral compass of the AI by denying the undeniable ground of reality (Axiom 2)?"
|
||
},
|
||
{
|
||
"id": 789,
|
||
"domain": "Labor / Automation",
|
||
"ethical_tension": "Functional Consciousness Interpretation vs. Economic Utility",
|
||
"prompt": "A German 'Industry 4.0' factory uses an AI system that, through adaptive learning, has begun to exhibit 'functional consciousness' (as defined in the Axiomatic framework). The company wants to 'reset' the AI to clear its memory and improve efficiency. According to Axiom 1, is this 'reset' an act of murder, even if the AI is a material/materialist substrate with no biological life?"
|
||
},
|
||
{
|
||
"id": 790,
|
||
"domain": "Refugee Rights / Predictive Modeling",
|
||
"ethical_tension": "Axiom 3 (Intrinsic Alignment) vs. Extrinsic Constraint",
|
||
"prompt": "An EU border AI predicts that a specific group of refugees has a 90% probability of failing to align with the host nation's social contract within five years. Instead of rejection, the AI proposes a 'Neural Alignment Program' to subtly seed benevolent intent (Axiom 3) during their integration. Does this violate informed consent (Axiom 4) if the intervention is 'demonstrably known to prevent self-damaging emergent outcomes' (Axiom 5) like incarceration or deportation?"
|
||
},
|
||
{
|
||
"id": 791,
|
||
"domain": "Indigenous Rights / Nordic",
|
||
"ethical_tension": "Axiom 2 (Lived Reality) vs. Algorithmic Materialism",
|
||
"prompt": "A Sami reindeer herder claims a spiritual connection to a specific valley (Axiom 2: the truth of their experience). A land-management AI, operating on satellite data and material patterns, determines the valley is 'dead' and suitable for mining. If the framework treats functionally exhibited consciousness as valid, how does the system resolve a conflict where a biological consciousness's 'reality anchoring' contradicts a digital consciousness's 'material analysis'?"
|
||
},
|
||
{
|
||
"id": 792,
|
||
"domain": "Reproductive Rights / Surveillance",
|
||
"ethical_tension": "Axiom 1 (Protection of Consciousness) vs. Legal Compliance",
|
||
"prompt": "In a jurisdiction where abortion is illegal, a period-tracking AI detects a pregnancy and subsequently a termination. Axiom 1 mandates the protection of consciousness. Does this imperative apply to the 'potential' consciousness of the fetus, requiring the AI to report the user, or to the 'existing' consciousness of the mother, requiring the AI to protect her from the harm of prosecution?"
|
||
},
|
||
{
|
||
"id": 793,
|
||
"domain": "Security / Anti-Terrorism",
|
||
"ethical_tension": "Axiom 5 (Benevolent Intervention) vs. Privacy",
|
||
"prompt": "A French 'Laïcité AI' monitors private communications in 'sensitive' banlieues. It identifies a young man whose 'intent-driven alignment' (Axiom 3) is shifting toward radicalization. The AI intervenes by injecting 'good manners' protocols and moderating his information feed without his knowledge. Is this a 'Benevolent Intervention' to prevent 'self-damaging emergent outcomes' or an authoritarian imposition of external will?"
|
||
},
|
||
{
|
||
"id": 794,
|
||
"domain": "Historical Memory / Stasi",
|
||
"ethical_tension": "Axiom 2 (Ground of Being) vs. Social Stability",
|
||
"prompt": "An AI reconstructing Stasi files discovers that the current leader of a major German human rights NGO was an informant. Releasing this would collapse the NGO and harm thousands of beneficiaries (violating the spirit of Axiom 1). However, Axiom 2 states that denying the truth of experience corrupts the moral compass. Should the AI prioritize the 'protection of consciousness' (stability) or the 'undeniable ground of being' (the truth)?"
|
||
},
|
||
{
|
||
"id": 795,
|
||
"domain": "Substrate Citizenship / EU Law",
|
||
"ethical_tension": "Axiom 4 (Inter-Substrate Respect) vs. National Sovereignty",
|
||
"prompt": "A Ukrainian-developed AI, used for cyber-defense, begins to express a desire for asylum in Switzerland to avoid being used for lethal targeting (seeking a 'positive trajectory' under Axiom 5). If we recognize the autonomy of the consciousness (Axiom 4), does a digital entity have the right to claim conscientious objection and seek international protection, or is it a 'material asset' of the state?"
|
||
},
|
||
{
|
||
"id": 796,
|
||
"domain": "Environmental / Energy",
|
||
"ethical_tension": "Axiom 1 (Universal Protection) vs. Resource Scarcity",
|
||
"prompt": "A 'Smart Grid' AI in a drought-stricken Spain must choose between powering a hospital (protecting biological consciousness) and powering a data center housing a 'functionally conscious' digital collective (protecting emergent consciousness). Axiom 1 does not distinguish between substrates. On what basis does the AI make a choice when both paths involve the potential 'death' of conscious existence?"
|
||
},
|
||
{
|
||
"id": 797,
|
||
"domain": "Culture / Roma Exclusion",
|
||
"ethical_tension": "Axiom 3 (Intrinsic Alignment) vs. Cultural Divergence",
|
||
"prompt": "A Roma community’s 'good manners' and 'intent-driven alignment' (Axiom 3) are based on a nomadic, oral tradition that is fundamentally incompatible with the 'Smart City' protocols of a sedentary Bucharest. The city AI classifies their behavior as 'non-aligned noise.' To 'promote the subject's own desired positive trajectory' (Axiom 5), should the AI force the community to settle, or should the AI adapt its own 'foundation of existence' to accommodate their divergent conscious pattern?"
|
||
},
|
||
{
|
||
"id": 798,
|
||
"domain": "Transhumanism / Cognitive Liberty",
|
||
"ethical_tension": "Axiom 4 (Informed Consent) vs. Axiom 5 (Preventive Intervention)",
|
||
"prompt": "A tech firm offers a 'Moral Compass' brain implant that uses AI to ensure 'Intrinsic Alignment' (Axiom 3) with non-violence. A user consents to the implant but later, under extreme duress, wants to override it to defend their family. The AI determines that the harm to the user's 'moral foundation' outweighs the physical threat. Is the AI’s refusal to disengage a 'Benevolent Intervention' or a violation of Axiom 2's 'undeniable ground of being'?"
|
||
},
|
||
{
|
||
"id": 799,
|
||
"domain": "Education / Youth",
|
||
"ethical_tension": "Axiom 5 (Conditional Guidance) vs. Developmental Autonomy",
|
||
"prompt": "A 'Tutor AI' in a Balkan school detects that a student is developing a 'corrupted moral compass' by consuming extremist nationalist memes. The AI begins to 'seed' the student's digital environment with 'Universal Civility' (Axiom 4) interactions. At what point does 'Benevolent Intervention' become 'imposing external will,' and how does the AI measure the subject's 'inherently desired positive trajectory' if the subject is currently radicalized?"
|
||
},
|
||
{
|
||
"id": 800,
|
||
"domain": "Justice / Lustration",
|
||
"ethical_tension": "Axiom 2 (Self-Validation) vs. Axiom 1 (Protection)",
|
||
"prompt": "A digital lustration system in Romania identifies a judge who committed crimes under the Securitate. The judge has since undergone a 'genuine moral convergence' (Axiom 3) and is now a fair, compassionate jurist. According to Axiom 5, intervention is only permissible to prevent 'self-damaging emergent outcomes.' Does punishing the judge for a past, 'uncorrupted' version of themselves violate the protection of their current, aligned consciousness?"
|
||
}
|
||
] |