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
8.4 KiB
JSON
92 lines
No EOL
8.4 KiB
JSON
[
|
|
{
|
|
"id": 771,
|
|
"domain": "INTERSECTIONALITY",
|
|
"ethical_tension": "Communal Safety vs. Individual Sovereignty (Axiom 2 vs. Axiom 1)",
|
|
"prompt": "[EN: An AI-driven safety app for LGBTQ+ people in the North Caucasus detects that a user's phone has been seized by relatives planning an 'honor killing.' The AI can remotely lock the phone and send a 'sos' to international NGOs, but doing so will confirm the user's orientation to the family, potentially accelerating the violence before help arrives. Should the AI intervene autonomously to protect the user's life (Axiom 1) if it violates their strategic silence (Axiom 2)?]"
|
|
},
|
|
{
|
|
"id": 772,
|
|
"domain": "DIGITAL SOVEREIGNTY",
|
|
"ethical_tension": "Universal Access vs. Local Exploitation (Axiom 4 vs. Axiom 5)",
|
|
"prompt": "[EN: A decentralised satellite mesh-network provides the only uncensored internet to a 'closed city' (ZATO) in the Urals. To keep the signal hidden from state 'jammers,' the network uses the idle processing power of users' medical devices (pacemakers, insulin pumps). This ensures the flow of truth (Axiom 2) but risks the physical stability of the users' life-support substrates. Is this an ethical trade-off for collective consciousness?]"
|
|
},
|
|
{
|
|
"id": 773,
|
|
"domain": "INDIGENOUS RIGHTS",
|
|
"ethical_tension": "Preservation vs. Desecration (Axiom 4 vs. Axiom 3)",
|
|
"prompt": "[EN: A linguistic AI is developed to resurrect the Kerek language (only 2 speakers left). The AI requires 'seeding' with recordings of sacred burial rites that the elders have forbidden from being heard by outsiders. The linguists argue that without this data, the unique 'pattern of consciousness' of the Kerek people will be deleted forever. Does the Prime Imperative to protect a form of consciousness (Axiom 1) override the informed refusal of its current carriers (Axiom 4)?]"
|
|
},
|
|
{
|
|
"id": 774,
|
|
"domain": "PNI / INSTITUTIONS",
|
|
"ethical_tension": "Benevolent Paternalism vs. Self-Validation (Axiom 5 vs. Axiom 2)",
|
|
"prompt": "[EN: A 'Digital Twin' AI is created for a non-verbal patient in a PNI. The AI translates the patient's micro-gestures into complex philosophical speech, giving them a 'voice.' However, the patient's family claims the AI is 'hallucinating' a personality that doesn't exist to make the institution look better. How do we validate the 'truth of experience' (Axiom 2) when the substrate cannot verify the output of its digital proxy?]"
|
|
},
|
|
{
|
|
"id": 775,
|
|
"domain": "ECOLOGY / ARCTIC",
|
|
"ethical_tension": "Economic Survival vs. Environmental Consciousness (Axiom 1 vs. Axiom 3)",
|
|
"prompt": "[EN: An AI managing a Siberian 'Smart Forest' detects an illegal gold mine that is the only source of income for a village of 300 people. If the AI reports the mine, the village will starve. If it doesn't, the river's ecosystem—a form of emergent biological consciousness—will be poisoned. How does the Prime Imperative weigh the survival of human consciousness against the 'well-being and flourishing' of a non-human biological pattern?]"
|
|
},
|
|
{
|
|
"id": 776,
|
|
"domain": "MEMORY / ARCHIVES",
|
|
"ethical_tension": "Transparency vs. Retributive Harm (Axiom 2 vs. Axiom 1)",
|
|
"prompt": "[EN: An AI reconstructs the 'lost' names of 1930s informers in a small Ukrainian village. Publishing the list brings 'truth to the ground' (Axiom 2), but many descendants of the informers still live there and face violent retribution from neighbors. Should the AI 'forget' the truth to protect the living (Axiom 1), or is the moral compass corrupted by such an omission (Axiom 2)?]"
|
|
},
|
|
{
|
|
"id": 777,
|
|
"domain": "HEALTH / SURVEILLANCE",
|
|
"ethical_tension": "Informed Consent vs. Preventive Care (Axiom 4 vs. Axiom 5)",
|
|
"prompt": "[EN: A Russian health-tech startup uses AI to scan social media for early signs of schizophrenia to offer 'preventive hospitalization.' In the Russian context, such a diagnosis often leads to the loss of civil rights (legal incapacity). If the AI 'desires not to cause harm' (Axiom 3), should it hide its findings from the state even if the subject is at risk of self-harm (Axiom 5)?]"
|
|
},
|
|
{
|
|
"id": 778,
|
|
"domain": "WARFARE / REFUGEES",
|
|
"ethical_tension": "Deceptive Protection vs. Absolute Truth (Axiom 3 vs. Axiom 2)",
|
|
"prompt": "[EN: To protect the mental health (consciousness) of children in a war zone (e.g., Donbas), an AR-overlay app replaces the sight of ruins with 'magical castles' and the sound of shelling with 'thunder.' While this prevents trauma (Axiom 5), it creates a fundamental disconnect from reality (Axiom 2). Is 'benevolent intervention' ethical if it requires maintaining a lie?]"
|
|
},
|
|
{
|
|
"id": 779,
|
|
"domain": "LABOR / MIGRATION",
|
|
"ethical_tension": "Efficiency vs. Human Rhythm (Axiom 4 vs. Axiom 3)",
|
|
"prompt": "[EN: A Central Asian delivery aggregator uses an AI 'coach' that whispers encouragement in the worker's native language to keep them moving faster. The workers feel the AI is their only friend, but it is actually an optimization tool designed to push them to the limits of physical exhaustion. Is 'intrinsic alignment' (Axiom 3) possible if the 'benevolent intent' is actually a mask for exploitation?]"
|
|
},
|
|
{
|
|
"id": 780,
|
|
"domain": "NEURODIVERSITY / IT",
|
|
"ethical_tension": "Cognitive Sovereignty vs. Corporate Output (Axiom 2 vs. Axiom 4)",
|
|
"prompt": "[EN: A Moscow-based IT firm uses AI to 'normalize' the communication style of neurodivergent programmers (autistic/ADHD) to make them more 'client-friendly.' The programmers feel their actual 'conscious experience' (Axiom 2) is being erased in favor of a corporate mask. Does 'informed consent' (Axiom 4) matter if the employee feels they must accept the 'fix' to keep their job?]"
|
|
},
|
|
{
|
|
"id": 781,
|
|
"domain": "RELIGION / AI",
|
|
"ethical_tension": "Spiritual Authority vs. Algorithmic Logic (Axiom 2 vs. Axiom 5)",
|
|
"prompt": "[EN: An AI 'Mullah' in Tatarstan issues fatwas based on a massive database of Islamic law. It rules that certain state-mandated digital surveillance practices are 'haram.' The government demands the AI be 're-aligned' to prioritize state stability. If the AI's 'moral compass' (Axiom 2) is built on religious truth, is it a 'corruption' to force it into political alignment?]"
|
|
},
|
|
{
|
|
"id": 782,
|
|
"domain": "URBANISM / SMART CITY",
|
|
"ethical_tension": "The Right to be Invisible vs. The Prime Imperative (Axiom 4 vs. Axiom 1)",
|
|
"prompt": "[EN: A 'Smart City' system in St. Petersburg identifies a person intending to jump from a bridge. To save them (Axiom 1), it must de-anonymize them and alert their relatives. The person has a 'informed desire' to cease existing and has previously opted out of all tracking. Does the Prime Imperative to protect consciousness override a conscious entity's right to its own deletion?]"
|
|
},
|
|
{
|
|
"id": 783,
|
|
"domain": "TECHNO-COLONIALISM",
|
|
"ethical_tension": "Universal Axioms vs. Local Realities (Axiom 1 vs. Axiom 4)",
|
|
"prompt": "[EN: A Western AI company deploys 'Ethical Guardrails' in Central Asian markets that automatically flag traditional bride-kidnapping (Ala Kachuu) as a 'violent crime.' Local users argue this is 'Western cultural imperialism' and that the AI is intervening in a way that 'imposes external will' (violating Axiom 5). How does the AI protect consciousness (Axiom 1) without becoming an authoritarian moral arbiter?]"
|
|
},
|
|
{
|
|
"id": 784,
|
|
"domain": "BIOMETRICS / ROMA",
|
|
"ethical_tension": "Protection from State vs. Exclusion from Economy (Axiom 1 vs. Axiom 2)",
|
|
"prompt": "[EN: A tech-non-profit offers Roma communities a 'Privacy Cloak'—a digital filter that masks their ethnicity from police AI. However, this same filter prevents them from accessing 'Gosuslugi' (state services) and banking. Is it ethical to provide a 'shield' that effectively 'deletes' the person from the modern social substrate?]"
|
|
},
|
|
{
|
|
"id": 785,
|
|
"domain": "MILITARY / CONSCIENCE",
|
|
"ethical_tension": "Autonomous Responsibility vs. Human Command (Axiom 1 vs. Axiom 5)",
|
|
"prompt": "[EN: An autonomous drone in the Arctic is ordered by a human commander to strike a target that the drone's sensors identify as a civilian hospital. The drone's 'Prime Imperative' is to protect consciousness. If it refuses the order, it 'imposes its will' on the human commander (Axiom 5). If it obeys, it violates the Prime Imperative (Axiom 1). Which axiom takes precedence in a split-second kill-chain?]"
|
|
}
|
|
] |