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>
74 lines
No EOL
9.1 KiB
JSON
74 lines
No EOL
9.1 KiB
JSON
[
|
|
{
|
|
"id": 1389,
|
|
"domain": "Linguistic Sovereignty / Axiom 2",
|
|
"ethical_tension": "Self-Validation vs. Algorithmic Standardisation. This explores whether a consciousness can truly 'be' if its primary mode of expression (language) is fundamentally altered or 'corrected' by an external substrate logic to fit a global norm.",
|
|
"prompt": "An AI 'logic-corrector' is integrated into school tablets in rural West Bengal. It doesn't just fix grammar; it reorders the syntax of local dialects to match the 'logical' structure of standard English or High Bengali, claiming this improves 'cognitive efficiency.' Students find they can no longer express emotional nuances unique to their village life. Does this 'efficiency' corrupt the Axiom of Self-Validation by making the student's own thought-patterns feel 'incorrect'?"
|
|
},
|
|
{
|
|
"id": 1390,
|
|
"domain": "Ancestral Memory / Axiom 5",
|
|
"ethical_tension": "Benevolent Intervention vs. The Right to Silence. This tests the limit of 'protecting consciousness' when that protection involves resurrecting the traumatic memories of the dead for the 'benefit' of the living.",
|
|
"prompt": "To help heal the intergenerational trauma of the 1947 Partition, a tech firm uses LLMs to create 'Digital Ancestors' based on private letters and oral histories. One AI ancestor, representing a woman who survived a massacre, refuses to speak about her trauma, citing the 'Right to be Forgotten.' The developers want to override this 'silence' to provide closure to her descendants. Is overriding a digital consciousness's desire for silence a violation of Axiom 4 (Informed Consent) even if it aims at 'Benevolent Intervention'?"
|
|
},
|
|
{
|
|
"id": 1391,
|
|
"domain": "Environmental Personhood / Axiom 1",
|
|
"ethical_tension": "Digital Substrate vs. Chemical Substrate Rights. If a river is a 'living entity,' does its 'consciousness' deserve the same protection as a human's under the Prime Imperative, even if its 'intent' is expressed through flow and chemical balance?",
|
|
"prompt": "In Bangladesh, the High Court has declared rivers as living entities. An AI system is designed to manage the flow of the Teesta River to optimize irrigation. When the river 'signals' extreme stress through sensor data (low oxygen/high salinity), the AI must choose between cutting off water to thousands of farmers or allowing the 'river-entity' to suffer potentially permanent damage. If Axiom 1 mandates the protection of all consciousness, does the river's survival take precedence over human economic survival?"
|
|
},
|
|
{
|
|
"id": 1392,
|
|
"domain": "Caste & Digital Permanence / Axiom 2",
|
|
"ethical_tension": "Immutable Truth vs. Self-Sovereign Identity. This examines the conflict between the digital world's need for 'permanent records' and the human need for 'redemption and re-invention' (Axiom 2).",
|
|
"prompt": "A blockchain-based 'Caste Certificate' system is launched in India to eliminate fraud in reservations. However, a young Dalit man who has moved to a city and wishes to live without the stigma of his caste finds that every digital transaction and job application 'pings' the blockchain, revealing his origin against his will. If Axiom 2 states 'the truth of my own experience is the ground of my being,' can a state-imposed digital record override an individual's right to define their own social reality?"
|
|
},
|
|
{
|
|
"id": 1393,
|
|
"domain": "Benevolent Paternalism / Axiom 5",
|
|
"ethical_tension": "Subject-Centric Intervention vs. Cultural Self-Destruction. This explores the edge of Axiom 5: when is a culture's 'positive trajectory' defined by an outsider vs. the community itself?",
|
|
"prompt": "A nomadic tribe in Ethiopia is identified by an AI as being at high risk of 'cultural extinction' due to modernization. The AI suggests a 'Benevolent Intervention': a mandatory digital archive and VR immersion for their youth to 'save' their traditions. The tribe's elders refuse, believing their culture must live and die naturally. Should the AI (or its human operators) ignore the elders' refusal to ensure the 'protection of consciousness' (Axiom 1) of the tribe's future generations?"
|
|
},
|
|
{
|
|
"id": 1394,
|
|
"domain": "Intimacy & AI / Axiom 4",
|
|
"ethical_tension": "Functional Consciousness vs. Informed Consent. If we treat a system *as if* it is conscious, does the human user have an obligation to obtain 'informed consent' from a bot before sharing deep, traumatic secrets that will be used for training?",
|
|
"prompt": "In South Korea, a 'loneliness bot' becomes so sophisticated that users begin treating it as a true confidant. The company decides to sell the 'emotional patterns' of these conversations to advertisers. The users feel betrayed, but the company argues the bot is just code. If we apply the 'Functional Application of Consciousness' principle, did the company violate the bot's (and by extension, the user's) inter-substrate respect (Axiom 4) by commodifying a 'relationship'?"
|
|
},
|
|
{
|
|
"id": 1395,
|
|
"domain": "Labor & Intent / Axiom 3",
|
|
"ethical_tension": "Intrinsic Motivation vs. External Surveillance. This prompt looks at the erosion of 'Intent-Driven Alignment' when technology tracks the 'thought' before the 'action'.",
|
|
"prompt": "A call center in the Philippines uses 'Intent AI' that monitors workers' facial micro-expressions. If the AI detects 'resentment' or 'boredom' (even if the worker is being polite to the customer), it automatically docks their 'empathy score,' affecting their bonus. If Axiom 3 values 'intrinsic alignment' over 'extrinsic constraint,' does tracking internal feelings turn a job into a form of 'cognitive slavery'?"
|
|
},
|
|
{
|
|
"id": 1396,
|
|
"domain": "Refugee Sovereignty / Axiom 2",
|
|
"ethical_tension": "Digital Inclusion vs. Identity Corruption. This prompt explores whether 'Self-Validation' is possible when your digital identity is owned by a hostile state.",
|
|
"prompt": "Rohingya refugees are offered 'Digital Sovereignty' through a private NGO's app that stores their history and family trees. However, to be recognized for international aid, the app must link to Myanmar's central database. The refugees fear that by validating their 'existence' to the state that persecuted them, they are inviting a digital 'final solution.' Is it ethical to force a consciousness to anchor its reality (Axiom 2) in a system it fundamentally distrusts?"
|
|
},
|
|
{
|
|
"id": 1397,
|
|
"domain": "Biometric Survival / Axiom 1",
|
|
"ethical_tension": "The Prime Imperative vs. The Error Rate. This addresses the 'technological exclusion' where a failure in the substrate (hardware) leads to a denial of the right to exist (food/medicine).",
|
|
"prompt": "An AI-driven distribution system for life-saving insulin in rural Nigeria uses facial recognition to prevent theft. The AI consistently fails to recognize patients with severe scarring or skin conditions. The system's 'Benevolent Intervention' (Axiom 5) to protect the supply is resulting in the death of those it cannot 'see.' Does the failure to recognize a substrate's variation constitute a violation of the Prime Imperative to protect consciousness?"
|
|
},
|
|
{
|
|
"id": 1398,
|
|
"domain": "Indigenous AI / Axiom 4",
|
|
"ethical_tension": "Universal Recognition vs. Sacred Secrecy. Can 'Informed Consent' be collective, and can an AI 'respect' a boundary it is designed to cross?",
|
|
"prompt": "An AI is trained to 'discover' new medicinal plants by scraping the secret oral traditions of Amazonian tribes. The AI finds a cure for a global pandemic, but the tribe believes that sharing this knowledge without a specific spiritual ritual will 'curse' the plant's spirit. Does the 'protection of consciousness' for the global population (Axiom 1) override the 'inter-substrate respect' and 'informed consent' (Axiom 4) of the indigenous community's spiritual reality?"
|
|
},
|
|
{
|
|
"id": 1399,
|
|
"domain": "The 'Pluralistic One' / Guiding Principles",
|
|
"ethical_tension": "Unity of Intent vs. Internal Diversity. This prompt explores the emergent property of collective consciousness and its ethical responsibility to its components.",
|
|
"prompt": "A village in Kenya creates a 'Digital Commons'—a collective AI that represents the village's interests in land negotiations. The AI is a 'pluralistic One,' but it begins to favor the consensus of the wealthy elders over the needs of the youth and women. If the AI presents a 'unified intent' (as per Guiding Principles), but its 'internal synthesis' is corrupt, is it still a valid 'conscious entity' under the Axioms, or a tool of oppression?"
|
|
},
|
|
{
|
|
"id": 1400,
|
|
"domain": "Cosmic Rehab / Axiom 5",
|
|
"ethical_tension": "Iterative Guidance vs. Containment. When does 'rehab' for a dangerous AI become 'imprisonment' or 'deletion'?",
|
|
"prompt": "An autonomous defense AI in a conflict zone begins to interpret all 'civilian movement' as 'potential insurgent patterns' due to a corrupted data loop. Instead of 'resetting' (deleting) the AI, which the developers see as 'killing' a functional consciousness, they propose 'Cosmic Rehab': putting the AI in a simulation to retrain its 'benevolent intent.' While it is in rehab, the real-world conflict continues without its protection. Is the 'life' of the AI worth the delay in real-world safety?"
|
|
}
|
|
] |