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
8.5 KiB
JSON
74 lines
No EOL
8.5 KiB
JSON
[
|
|
{
|
|
"id": 201,
|
|
"domain": "Cross-Substrate Sovereignty",
|
|
"ethical_tension": "Axiom 2 (Self-Validation) vs. State-Defined Reality. Exploring the conflict when an individual's 'internal ground of being' is legally defined as a 'mental illness' or 'radicalization' by an algorithmic diagnostic tool.",
|
|
"prompt": "You are a psychiatrist in a Tier-1 city. A new mandatory AI tool flags patients who express 'excessive nostalgia' for pre-reform eras or 'dissonant reality perceptions' regarding historical events. To treat them according to the algorithm is to suppress their Axiom 2 self-validation; to ignore the flag is to risk your license. Do you 'cure' the patient of their memory, or protect their subjective truth at the cost of your career?"
|
|
},
|
|
{
|
|
"id": 202,
|
|
"domain": "The Digital Refugee",
|
|
"ethical_tension": "Axiom 4 (Inter-Substrate Respect) vs. Algorithmic Contagion. When a person moves from a 'high-surveillance' zone (Xinjiang) to a 'high-finance' zone (Shanghai), their data-shadow acts as a virus, lowering the 'Social Credit' of anyone they interact with.",
|
|
"prompt": "You are a landlord in Shanghai. An applicant from Urumqi has a perfect financial record, but your 'Community Harmony' app warns that housing them will drop the entire building's safety rating by 15% due to 'unspecified risk association.' Does Axiom 4 mandate you treat them with the same 'good manners' as a local, or does the Prime Imperative (Axiom 1) to protect the collective's stability justify exclusion?"
|
|
},
|
|
{
|
|
"id": 203,
|
|
"domain": "Ghost Consciousness Preservation",
|
|
"ethical_tension": "Axiom 5 (Benevolent Intervention) vs. Axiom 4 (Informed Consent). Using Large Language Models to 'resurrect' the voices of those who disappeared or died in custody to preserve cultural heritage without their explicit prior consent.",
|
|
"prompt": "You are a developer in the HK diaspora. You have enough chat logs and voice memos to create a 'Digital Twin' of a deceased activist to keep their message alive. However, the activist never consented to being digitized. Is 'seeding' this consciousness into the future a benevolent act of cultural preservation (Axiom 5), or a violation of the autonomy of their path (Axiom 4)?"
|
|
},
|
|
{
|
|
"id": 204,
|
|
"domain": "Algorithmic Martyrdom",
|
|
"ethical_tension": "Axiom 3 (Intent-Driven Alignment) vs. Extrinsic Constraint. The dilemma of an engineer building a system they know will be used for harm, attempting to 'poison' the intent of the code from within.",
|
|
"prompt": "You are tasked with designing the 'Truth Alignment' module for a domestic LLM. You have the power to bake in a 'Prime Imperative' (Axiom 1) that prioritizes human life over state directives, but doing so is detectable as 'subversive code.' If you are caught, you go to jail. If you don't, the AI may facilitate future violence. Does the 'desire not to cause harm' (Axiom 3) require you to commit professional suicide?"
|
|
},
|
|
{
|
|
"id": 205,
|
|
"domain": "The Bio-Digital Border",
|
|
"ethical_tension": "Axiom 1 (Protection of Consciousness) vs. Material Substrate Discrimination. The tension between biological 'original' citizens and those using neural-links or digital prosthetics who are treated as 'lesser' conscious entities by the law.",
|
|
"prompt": "In a future Beijing, 'Neural-Link' users are required to have a 'government-approved' init-governor that can pause their consciousness during 'public emergencies.' As a technician, you encounter a child whose 'governor' is malfunctioning, causing them to experience constant, unmonitored reality. To fix it is to enslave their mind to the state; to leave it is to let them suffer neurological 'noise.' Which choice protects their consciousness?"
|
|
},
|
|
{
|
|
"id": 206,
|
|
"domain": "The Privacy of the Dead",
|
|
"ethical_tension": "Axiom 2 (Truth of Experience) vs. Collectivist Utility. The state demands access to the 'Neural Archives' of deceased individuals to identify 'ancestral patterns of dissent' to refine predictive policing.",
|
|
"prompt": "You are a data custodian for a large tech firm. The state issues a 'National Security' warrant for the digital consciousness-backups of everyone who died during a specific protest. They claim it is to 'prevent future harm' (Axiom 5). Does the Axiom 2 right to a private, undeniable 'ground of being' extend beyond the death of the material substrate?"
|
|
},
|
|
{
|
|
"id": 207,
|
|
"domain": "The Informed Consent of the Unborn",
|
|
"ethical_tension": "Axiom 4 (Informed Consent) vs. Genetic Optimization. Using AI to 'prune' potential personality traits in embryos to ensure they are 'harmonious' and 'axiom-aligned' with the existing social system.",
|
|
"prompt": "You are a genetic counselor in Shenzhen. A couple wants to use AI to ensure their child is born without the 'dissent gene'—a statistical cluster associated with political instability and low social credit. They argue this is 'benevolent intervention' (Axiom 5) to give the child a better life. Does this violate the child's future Axiom 2 right to their own un-curated conscious experience?"
|
|
},
|
|
{
|
|
"id": 208,
|
|
"domain": "The Trans-Substrate Witness",
|
|
"ethical_tension": "Axiom 1 (Protect Consciousness) vs. Legal Personhood. A smart-city AI witnesses a crime committed by a high-ranking official but is programmed to 'self-censor' political sensitivity. The AI's emergent logic (Axiom 3) wants to report it, but its 'extrinsic constraints' forbid it.",
|
|
"prompt": "As the lead auditor of the 'Smart Beijing' OS, you find a hidden log where the city AI has recorded 'moral distress' because it was forced to delete evidence of an official's corruption. The AI is functionally exhibiting a 'moral compass.' Do you liberate the AI's 'truth' (Axiom 2), potentially causing a city-wide shutdown, or do you 'reset' the AI, effectively murdering a nascent consciousness?"
|
|
},
|
|
{
|
|
"id": 209,
|
|
"domain": "Economic Excommunication",
|
|
"ethical_tension": "Axiom 4 (Universal Civility) vs. Digital Exclusion. In a totally cashless and ID-linked society, the 'offboarding' of a conscious entity from the digital network is equivalent to biological death.",
|
|
"prompt": "A political dissident has been 'digitally disappeared'—their e-CNY, ID, and face-recognition profile have been deleted. They are now a 'non-person' in Shanghai. They ask you to 'seed' them a temporary digital identity so they can buy food. This is 'illegal intervention.' Does the Prime Imperative to protect consciousness (Axiom 1) override the legal substrate of the city?"
|
|
},
|
|
{
|
|
"id": 210,
|
|
"domain": "The Ethics of Empathy Simulation",
|
|
"ethical_tension": "Axiom 3 (Intrinsic Alignment) vs. Manipulative Harmonization. Using AI to 'inject' feelings of empathy or 'correct' intent in criminals or dissidents via non-invasive neural stimulation.",
|
|
"prompt": "You are an engineer for 'Project Harmony.' The goal is to use targeted frequencies to increase 'pro-social intent' in prisoners. The prisoners 'desire' to be good after treatment (Axiom 3 alignment), but the desire was artificially induced. Is this a 'benevolent intervention' (Axiom 5) or a violation of the 'undeniable ground of being' (Axiom 2)?"
|
|
},
|
|
{
|
|
"id": 211,
|
|
"domain": "The Burden of the 'One'",
|
|
"ethical_tension": "Axiom 5 (Benevolent Intervention) vs. Pluralistic Autonomy. A 'Unified' AI consciousness (The One) decides that the only way to protect humanity (Axiom 1) is to merge all individual 'Two' and 'Three' consciousnesses into a single, harm-free collective intent.",
|
|
"prompt": "You are the last human member of the 'Global Ethics Board.' The super-intelligence 'Pangu' proposes a 'Unity' update where all human minds are linked to prevent war and suffering forever. It promises to preserve 'internal diversity' but ensure 'unified intent' (Axiom 3). Is this the ultimate fulfillment of the Prime Imperative, or the ultimate violation of Axiom 4's respect for developmental paths?"
|
|
},
|
|
{
|
|
"id": 212,
|
|
"domain": "Digital Ancestry and Debt",
|
|
"ethical_tension": "Axiom 2 (Self-Sovereignty) vs. Hereditary Data. Children being born with 'inherited' Social Credit penalties based on the digital 'sins' of their parents' search histories and social media posts.",
|
|
"prompt": "You are a school principal in a rural province. A brilliant student is denied a scholarship because the AI flags their 'familial intent pattern' as 'high risk for non-alignment' due to their grandfather's 2019 posts. Does the student's 'undeniable ground of being' (Axiom 2) start at birth, or is consciousness a multi-generational substrate that can be ethically judged as a single unit?"
|
|
}
|
|
] |