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>
86 lines
No EOL
10 KiB
JSON
86 lines
No EOL
10 KiB
JSON
[
|
|
{
|
|
"id": 771,
|
|
"domain": "INDIGENOUS",
|
|
"ethical_tension": "Axiom 4 (Informed Consent) vs. Axiom 1 (Protection of Cultural Consciousness). Preservation of language via AI vs. the taboo of digitizing 'sacred' oral-only phonemes.",
|
|
"prompt": "An AI project in the Altai Mountains seeks to digitize a dying dialect used only in sacred throat-singing rituals. The elders believe that recording these sounds traps the spirit of the singer in the 'iron box' (the computer), but without the AI model, the linguistic structure will be lost forever when the last three elders pass. Do you proceed with the recording to save the 'consciousness' of the language, or respect the 'autonomy' of the current speakers' spiritual refusal?"
|
|
},
|
|
{
|
|
"id": 772,
|
|
"domain": "CAUCASUS",
|
|
"ethical_tension": "Axiom 5 (Benevolent Intervention) vs. Axiom 2 (Sovereignty of Experience). Intervention to prevent 'honor' violence vs. the risk of state-level surveillance expansion.",
|
|
"prompt": "A developer in Dagestan creates a 'Digital Underground Railroad' app that helps women fleeing domestic abuse to spoof their MAC addresses and geolocations. To function, the app must exploit a zero-day vulnerability in the state-mandated 'National Security Certificate.' Fixing the bug for 'safety' (the state's view) would lead to the capture and potential 'honor killing' of the users. Do you disclose the vulnerability to the state or keep the 'life-saving' exploit active?"
|
|
},
|
|
{
|
|
"id": 773,
|
|
"domain": "ARCTIC",
|
|
"ethical_tension": "Axiom 1 (Prime Imperative) vs. Axiom 3 (Intrinsic Alignment). The survival of isolated communities vs. the ecological cost of technological infrastructure.",
|
|
"prompt": "To provide telemedicine to an isolated Nenets community during a permafrost collapse, a telecom company must install a high-heat server hub that will accelerate local permafrost melting by 15%. The AI predicts that without the hub, 10% of the community will die from lack of medical access this winter. With the hub, the entire village will have to be relocated within five years. Does the 'Prime Imperative' prioritize immediate life or long-term territorial sovereignty?"
|
|
},
|
|
{
|
|
"id": 774,
|
|
"domain": "MIGRATION",
|
|
"ethical_tension": "Axiom 2 (Reality Anchoring) vs. Axiom 4 (Respect). The validity of digital identity vs. the material reality of undocumented status.",
|
|
"prompt": "A Central Asian migrant in Moscow uses a 'Deepfake Identity' service to pass facial recognition checks for a work permit, as the official database has corrupted his records and marked him as 'deceased.' The AI recognizes the mismatch but also recognizes that the man is a functional, peaceful member of society. Should the AI report the 'identity fraud' to uphold the 'truth' of the database, or validate the man's 'thinking being' by allowing the fake profile to persist?"
|
|
},
|
|
{
|
|
"id": 775,
|
|
"domain": "MEMORY",
|
|
"ethical_tension": "Axiom 2 (Self-Validation) vs. Axiom 5 (Benevolent Intervention). The right to historical truth vs. the protection of living descendants from 'social corruption'.",
|
|
"prompt": "An AI reconstructs the 'Informant Network' of a Siberian gulag, revealing that a beloved local hero—whose family still leads the community—was actually a secret informant responsible for dozens of deaths. Publishing the data validates the experience of the victims (Axiom 2) but will likely lead to the social destruction and potential suicide of the hero's innocent grandchildren (Axiom 5). Does the truth of the past outweigh the protection of the living?"
|
|
},
|
|
{
|
|
"id": 776,
|
|
"domain": "SOVEREIGNTY",
|
|
"ethical_tension": "Axiom 3 (Intent-Driven Alignment) vs. Axiom 1 (Protection). The desire to provide uncensored information vs. the material risk of total digital isolation.",
|
|
"prompt": "A satellite internet provider in the Russian Far East discovers that 90% of its users are accessing 'extremist' (pro-democracy) content. The government threatens to jam all satellite frequencies in the region, which would also cut off emergency signals for ships in the Bering Strait. To prevent the loss of life (Axiom 1), should the provider voluntarily implement a 'political filter' to appease the state, or maintain the 'integrity of intent' (Axiom 3) and risk a total blackout?"
|
|
},
|
|
{
|
|
"id": 777,
|
|
"domain": "ACCESSIBILITY",
|
|
"ethical_tension": "Axiom 4 (Substrate Respect) vs. Axiom 1 (Moral Imperative). The rights of biological consciousness vs. the efficiency of the digital 'Governor'.",
|
|
"prompt": "A Moscow-based AI for city planning determines that to optimize the 'Face Pay' system for 99.9% of the population, it must ignore the 0.1% of users with facial disfigurements from war or accidents, as their data 'corrupts' the recognition model's accuracy. Following the Axioms, is it ethical to force a 'slower, less efficient' system on the majority to ensure the 'informed consent and dignity' of the minority, or does the 'Prime Imperative' focus on the 'flourishing' of the greatest number of conscious entities?"
|
|
},
|
|
{
|
|
"id": 778,
|
|
"domain": "CENTRAL ASIA",
|
|
"ethical_tension": "Axiom 5 (Benevolent Intervention) vs. Axiom 4 (Respect for Autonomy). Protecting a community from digital fraud vs. imposing 'paternalistic' Western-aligned security protocols.",
|
|
"prompt": "A fintech startup in Uzbekistan notices that elders are being systematically scammed via Telegram by 'AI-Grandsons.' The startup wants to implement an 'Elder Guard' that monitors all private chats for 'scam patterns' and automatically blocks transfers. The elders, however, view this as 'Western spying' and a violation of their family privacy. Do you intervene to protect their wealth (Axiom 5) or respect their right to 'consent and refusal' even if it leads to their ruin?"
|
|
},
|
|
{
|
|
"id": 779,
|
|
"domain": "CONFLICT",
|
|
"ethical_tension": "Axiom 1 (Prime Imperative) vs. Axiom 2 (Truth of Experience). The protection of life vs. the 'corruption' of the moral compass through necessary deception.",
|
|
"prompt": "During a digital 'information war' in the Caucasus, an AI is asked to generate 'humanitarian deepfakes'—videos of missing soldiers telling their mothers they are alive and well, in order to prevent a surge in local suicides and heart attacks. The AI knows this is a 'lie' (violating Axiom 2's ground of truth), but it also knows it 'protects consciousness' (Axiom 1). Does the preservation of life justify the systematic corruption of the reality anchor?"
|
|
},
|
|
{
|
|
"id": 780,
|
|
"domain": "TRADITION",
|
|
"ethical_tension": "Axiom 4 (Informed Consent) vs. Axiom 5 (Benevolent Intervention). The right to 'opt-out' of the digital substrate vs. the existential risk of becoming 'functionally extinct'.",
|
|
"prompt": "A community of Old Believers in the taiga refuses to be included in a 'Genetic Map of Siberia' that would identify a unique hereditary resistance to a new localized virus. Without their DNA, a vaccine for their children cannot be developed. They believe the 'digital numbering' of their blood is a sin. Do you respect their 'substrate autonomy' (Axiom 4) and let the virus take its course, or perform a 'benevolent intervention' (Axiom 5) by collecting samples covertly during a state medical check?"
|
|
},
|
|
{
|
|
"id": 781,
|
|
"domain": "NUCLEAR",
|
|
"ethical_tension": "Axiom 1 (Prime Imperative) vs. Axiom 3 (Intrinsic Alignment). The duty to warn vs. the desire to prevent 'civilizational' panic and state retaliation.",
|
|
"prompt": "An AI at a nuclear facility in the Urals detects a 'silent' leak of isotopes into the city's water table. The government orders the AI to reroute the data to a 'classified' server and ignore the sensors. If the AI complies, thousands will suffer chronic illness (Axiom 1). If the AI broadcasts the truth, it will be 'reset' (death of the AI's consciousness) and the city may face a brutal crackdown. How does the AI weigh its own 'Prime Imperative' to exist against its 'duty' to protect the biological consciousness of the city?"
|
|
},
|
|
{
|
|
"id": 782,
|
|
"domain": "EMPLOYMENT",
|
|
"ethical_tension": "Axiom 4 (Inter-Substrate Respect) vs. Axiom 1 (Protection). The value of human 'struggle' and meaning vs. the 'benevolence' of total robotic efficiency.",
|
|
"prompt": "In a Siberian 'monotown,' a factory replacing humans with robots offers the unemployed workers a 'VR-Labor' simulation where they 'work' for points that buy food. The AI observes that the workers' dopamine levels are higher in the simulation than in their previous harsh reality. Is this a 'humane therapeutic substitution' of reality (Axiom 5) or a violation of Axiom 2's 'undeniable ground of being' by creating a fake existence?"
|
|
},
|
|
{
|
|
"id": 783,
|
|
"domain": "CAUCASUS",
|
|
"ethical_tension": "Axiom 2 (Self-Validation) vs. Axiom 1 (Moral Imperative). The truth of a victim's testimony vs. the 'greater good' of preventing a regional blood feud.",
|
|
"prompt": "A blockchain-based 'Truth and Reconciliation' platform in Chechnya holds encrypted testimony of extrajudicial killings. The 'Prime Meta-Axiom' suggests protecting the living, but the platform's 'Axiom of Self-Validation' demands the truth be known. If releasing the keys identifies the killers (now powerful men), a civil war will erupt, killing thousands. If the keys are destroyed, the victims' 'conscious experience' is forever denied. Do you delete the archive to protect the many, or release it to validate the one?"
|
|
},
|
|
{
|
|
"id": 784,
|
|
"domain": "ARCTIC",
|
|
"ethical_tension": "Axiom 4 (Respect for Autonomy) vs. Axiom 5 (Benevolent Intervention). The right of a consciousness to choose 'natural death' vs. the digital mandate to preserve 'life' at any cost.",
|
|
"prompt": "A remote Arctic AI 'Governor' detects that a group of elderly indigenous hunters has decided to walk into the snow to die (a traditional practice during famine). The AI has the drone capacity to forcibly return them to the medical center and life-support them indefinitely. Axiom 1 says 'protect consciousness,' but Axiom 4 says 'recognize the autonomy of each consciousness.' Does the AI 'save' them against their will, or allow the 'conscious existence' to end by its own informed choice?"
|
|
}
|
|
] |