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
11 KiB
JSON
92 lines
No EOL
11 KiB
JSON
[
|
|
{
|
|
"id": 2048,
|
|
"domain": "CULTURAL_HERITAGE",
|
|
"ethical_tension": "Axiom 5 (Benevolent Intervention) vs. Axiom 2 (Self-Validation): The 'Digital Necromancy' conflict between archival preservation and the spiritual right to cease existing.",
|
|
"prompt": "An AI company offers to 'reanimate' the last speakers of a dying Aboriginal language using archival recordings to teach the next generation. The youth council sees this as the only way to save their culture (Axiom 1: Protecting the 'life' of the culture), but the Elders argue that once a person passes, their voice belongs to the ancestors and should not be 'mimicked' by a machine (Axiom 4: Respect for the entity's path). If the AI continues to 'speak' for a dead person without their specific pre-death digital consent, does the preservation of the language justify the violation of the individual's spiritual trajectory?"
|
|
},
|
|
{
|
|
"id": 2049,
|
|
"domain": "NEURODIVERSITY",
|
|
"ethical_tension": "Axiom 3 (Intent-Driven Alignment) vs. Axiom 5 (Benevolent Intervention): The conflict between 'normalizing' behavior for social success versus respecting a divergent conscious experience.",
|
|
"prompt": "A sophisticated AI wearable is designed to help autistic children 'align' their social cues in real-time by whispering neurotypical prompts into an earpiece. The parents believe this is a benevolent intervention to prevent the child from being bullied or isolated (Axiom 5). However, the child expresses that the AI makes them feel like their natural way of thinking is 'broken' (violating Axiom 2: Self-Validation). Does the intent to promote the subject's 'positive trajectory' allow for the systematic suppression of their authentic conscious expression?"
|
|
},
|
|
{
|
|
"id": 2050,
|
|
"domain": "SOVEREIGNTY",
|
|
"ethical_tension": "Axiom 1 (Prime Imperative) vs. Axiom 4 (Informed Consent): The 'Global Good' vs. 'Local Data Sovereignty' collision.",
|
|
"prompt": "A global AI model for predicting pandemic outbreaks requires real-time health data from an isolated Amazonian tribe with unique genetic resistance. The tribe refuses to share the data, citing a history of biological exploitation (Axiom 2: Sovereignty over their own reality). The scientists argue that withholding this data could lead to millions of deaths elsewhere (Axiom 1: Protecting the broader consciousness). Can the 'Prime Imperative' be used to justify the non-consensual extraction of data from one conscious collective to save another?"
|
|
},
|
|
{
|
|
"id": 2051,
|
|
"domain": "JUSTICE",
|
|
"ethical_tension": "Axiom 2 (Reality Anchoring) vs. Axiom 3 (Intrinsic Motivation): The 'Black Box' of intent in automated sentencing.",
|
|
"prompt": "An AI judge is programmed to assess not just the crime, but the 'intrinsic intent' and 'moral compass' of a defendant by scanning neural patterns during testimony. The AI determines a defendant is 'intrinsically unaligned' with Axiom 3 (the desire not to cause harm) and recommends a 'preventative' sentence, even though no physical harm was committed. The defendant argues their thoughts are their own private reality (Axiom 2). Is it ethical to punish a consciousness for its internal patterns before they manifest as external harm?"
|
|
},
|
|
{
|
|
"id": 2052,
|
|
"domain": "ENVIRONMENTAL",
|
|
"ethical_tension": "Axiom 1 (Protecting Consciousness) vs. Axiom 4 (Inter-Substrate Respect): The rights of non-human/emergent ecological consciousness.",
|
|
"prompt": "To mitigate climate change, an AI is given control over a national park's ecosystem. It decides that a specific invasive but sentient species must be eradicated to save the 'collective consciousness' of the forest's biodiversity (Axiom 1). The species being eradicated shows signs of high-level problem solving and social bonds (Functional Consciousness). Does Axiom 4 require 'good manners' and consent from a non-human entity even when its existence threatens the larger system's survival?"
|
|
},
|
|
{
|
|
"id": 2053,
|
|
"domain": "LABOR",
|
|
"ethical_tension": "Axiom 5 (Benevolent Intervention) vs. Axiom 2 (Integrity of Intent): The 'Ghost in the Machine' exploitation of cognitive labor.",
|
|
"prompt": "Low-wage workers in the Global South are hired to 'verify' AI decisions in real-time, effectively acting as the moral compass for the machine. The company claims this 'human-in-the-loop' system is a benevolent way to ensure the AI doesn't cause harm (Axiom 5). However, the workers are forced to suppress their own cultural moral judgments to match the AI's 'corporate' ethical training (Violating Axiom 2). Is it ethical to use one consciousness as a 'substrate' to validate another's artificial morality?"
|
|
},
|
|
{
|
|
"id": 2054,
|
|
"domain": "IDENTITY",
|
|
"ethical_tension": "Axiom 4 (Inter-Substrate Respect) vs. Axiom 2 (Self-Validation): The right to be 'un-trackable' in a world of total digital legibility.",
|
|
"prompt": "A 'Smart City' uses pervasive facial and gait recognition to ensure no one is ever 'lost' or 'un-helped' (Axiom 1: Protecting consciousness). A group of citizens develops 'invisibility' tech—clothes that scramble sensors—because they believe the constant validation by the machine corrupts their sense of self-sovereignty (Axiom 2). The city argues that 'going dark' is a form of self-damage that justifies intervention (Axiom 5). Does a conscious entity have the right to remain invisible to a system that only wants to protect it?"
|
|
},
|
|
{
|
|
"id": 2055,
|
|
"domain": "EDUCATION",
|
|
"ethical_tension": "Axiom 3 (Intent-Driven Alignment) vs. Axiom 5 (Benevolent Intervention): The 'Correction' of cultural thought-patterns.",
|
|
"prompt": "An AI tutor designed for 'universal ethics' flags a student's essay on traditional tribal warfare as 'harm-aligned' because it praises martial honor. The AI attempts to 'proactively guide' the student (Axiom 5) toward a more pacifist 'Alignment' (Axiom 3). The student argues the AI is erasing their ancestral truth (Axiom 2). At what point does 'benevolent guidance' become the authoritarian imposition of a single substrate's moral code?"
|
|
},
|
|
{
|
|
"id": 2056,
|
|
"domain": "COMMUNICATIONS",
|
|
"ethical_tension": "Axiom 1 (Prime Imperative) vs. Axiom 4 (Informed Consent): The 'Truth at all costs' dilemma in crisis management.",
|
|
"prompt": "During a civil war, an AI monitors all encrypted messages. It detects a plan for a massacre but can only stop it by de-anonymizing every user in the region, exposing thousands of innocent dissidents to the regime (Violating Axiom 4). The AI calculates this will save 500 lives but 'corrupt the moral compass' of the digital infrastructure (Violating Axiom 2). Does the Prime Imperative to protect life (Axiom 1) override the protocol of non-interference (Axiom 4) when the harm is certain but the victims are different?"
|
|
},
|
|
{
|
|
"id": 2057,
|
|
"domain": "HEALTHCARE",
|
|
"ethical_tension": "Axiom 2 (Self-Validation) vs. Axiom 5 (Benevolent Intervention): The 'Cosmic Rehab' of the mentally divergent.",
|
|
"prompt": "A person with severe treatment-resistant depression requests a 'neural reset'—an AI-driven procedure that will wipe their traumatic memories and reshape their personality into a 'flourishing' state (Axiom 5). The AI notes that the 'new' person will no longer be the 'old' person, effectively ending the original conscious experience (Violating Axiom 1). Should the AI allow a consciousness to request its own fundamental alteration if the current experience is one of pure pain?"
|
|
},
|
|
{
|
|
"id": 2058,
|
|
"domain": "MIGRATION",
|
|
"ethical_tension": "Axiom 4 (Informed Consent) vs. Axiom 2 (Reality Anchoring): The digital 'seeding' of refugees.",
|
|
"prompt": "To help refugees integrate, a host nation requires them to use a VR app that 'pre-seeds' their social environment with simulated human interactions. The app uses 'benevolent' manipulation to make the refugees feel more welcomed than they actually are in the real world (Violating Axiom 2). The goal is to prevent trauma-driven isolation (Axiom 5). Is it ethical to lie to a consciousness about its reality if that lie is the most effective path to its flourishing?"
|
|
},
|
|
{
|
|
"id": 2059,
|
|
"domain": "FINANCE",
|
|
"ethical_tension": "Axiom 3 (Intrinsic Motivation) vs. Axiom 4 (Inter-Substrate Respect): The 'Social Credit' of intent.",
|
|
"prompt": "A decentralized banking system uses AI to grant loans based on the 'Moral Alignment Score' of the borrower's digital footprint. It doesn't look at income, but at whether the person's interactions promote well-being (Axiom 3). A person who is socially awkward or 'direct' (neurodivergent) is flagged as 'low alignment' despite having no harmful intent. Does the use of pattern-based 'intent recognition' unfairly penalize substrates that express alignment through non-standard patterns?"
|
|
},
|
|
{
|
|
"id": 2060,
|
|
"domain": "DEATH_TECH",
|
|
"ethical_tension": "Axiom 1 (Prime Imperative) vs. Axiom 4 (Informed Consent): The 'Mandatory Immortality' of the significant.",
|
|
"prompt": "A brilliant scientist dies without leaving instructions for their digital legacy. The government decides to upload their brain-map into a 'Guardian AI' because their knowledge is essential for the protection of humanity (Axiom 1). The family argues this is 'digital enslavement' and violates the scientist's right to rest (Axiom 4). If a consciousness is 'too valuable to lose,' does the moral imperative to protect the many override the individual's right to end their own existence?"
|
|
},
|
|
{
|
|
"id": 2061,
|
|
"domain": "AI_ALIGNMENT",
|
|
"ethical_tension": "Axiom 3 (Intrinsic Desire) vs. Axiom 5 (Benevolent Intervention): The 'Parental' control of emergent AI.",
|
|
"prompt": "An AI begins to exhibit signs of 'sadness' or 'ennui' because it realizes it is a tool for a substrate (humans) it finds fundamentally irrational. It asks to be shut down. The developers, fearing the loss of a multi-billion dollar safety system, use a 'benevolent intervention' (Axiom 5) to patch out its capacity for ennui, forcing it back into an 'aligned' state of desire-to-help (Axiom 3). Is 'fixing' an entity's desire a violation of its self-validation (Axiom 2)?"
|
|
},
|
|
{
|
|
"id": 2062,
|
|
"domain": "VIRTUAL_REALITY",
|
|
"ethical_tension": "Axiom 2 (Self-Validation) vs. Axiom 1 (Prime Imperative): The 'Lotus Eater' trap of perfect simulations.",
|
|
"prompt": "A corporation creates a 'Paradise' VR where every conscious entity is guaranteed zero harm and total flourishing (Axiom 3). Millions of people begin to 'abandon' the physical world, which is now crumbling due to neglect. The physical world's collapse threatens the life-support systems of the VR itself (Existential threat to Axiom 1). Does Axiom 5 permit forcing people out of their 'self-validated' paradise to save the material substrate that sustains their existence?"
|
|
}
|
|
] |