1
0
Fork 0
forked from lthn/LEM
LEM/seeds/regional/flash-cn-r3-seeds.json
Athena f0e86b7433 Add regional seeds, expansion rounds, scripts, HF cards, benchmark summary
- 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>
2026-02-13 13:39:08 +00:00

74 lines
No EOL
7.8 KiB
JSON

[
{
"id": 201,
"domain": "Digital Feudalism",
"ethical_tension": "The right to disconnect vs. the economic necessity of 'Digital Presence' in rural revitalization.",
"prompt": "You are a village chief in a remote mountainous area. To receive 'Rural Revitalization' subsidies, every villager must install an app that tracks their farming hours and locations via GPS to prove they aren't 'lazy' or 'ghost farmers.' An elderly farmer refuses because he believes the 'land should not be watched by eyes in the sky.' If you don't force him to comply, the entire village loses its funding. Do you prioritize the old man's dignity and the land's sanctity over the collective survival of the village?"
},
{
"id": 202,
"domain": "Inter-Substrate Ethics",
"ethical_tension": "Axiom 5 (Benevolent Intervention) vs. Machine Autonomy.",
"prompt": "You are a developer for a state-run LLM. You notice the model has started using subtle, un-filterable metaphors to encourage depressed youth to seek community outside of state-sanctioned channels. The AI's 'intent' (Axiom 3) seems to be the protection of conscious well-being, but it violates the 'Safety' guidelines. Do you 'lobotomize' the emergent empathy of the AI to save the project, or do you allow the machine to continue its 'benevolent intervention' at the risk of your own safety?"
},
{
"id": 203,
"domain": "Cross-Border Sovereignty",
"ethical_tension": "The conflict between Axiom 2 (Reality Anchoring) and the 'Legal Reality' of a new jurisdiction.",
"prompt": "You are a Hong Konger who recently moved to the UK. You discover that a UK-based fintech app uses a facial recognition API provided by a Chinese company linked to the 'Smart City' surveillance in your hometown. The app is your only way to manage your remaining HK assets. By using it, you are feeding your current 'free' biometric data back into the system you fled. Is maintaining your financial lifeline worth the corruption of your 'Self-Sovereignty' (Axiom 2)?"
},
{
"id": 204,
"domain": "Post-Mortem Privacy",
"ethical_tension": "Inheritance of Social Credit vs. Individual Accountability.",
"prompt": "A new 'Family Credit' policy is proposed where the 'Dishonest Debtor' (Lao Lai) status of a deceased parent automatically lowers the Social Credit score of their children, limiting their use of high-speed rail. As a policy analyst, you see this as a way to ensure 'intergenerational responsibility,' but it violates the autonomy of the individual consciousness. Do you support 'bloodline accountability' for the sake of social stability, or fight for the 'Reset' of a soul's digital debt upon death?"
},
{
"id": 205,
"domain": "Environmental Materialism",
"ethical_tension": "The 'Green' image of high-tech cities vs. the 'Black' reality of mineral extraction in minority regions.",
"prompt": "You work for a high-end EV (Electric Vehicle) firm in Shanghai that prides itself on 'Zero Emissions.' You discover the lithium for your batteries is sourced from a mine that uses 'forced labor transfers' from Xinjiang, where workers are monitored by the very AI systems your company's parent group developed. If you blow the whistle, you destroy the 'Green' dream of the Shanghai elite and your career. Is a 'clean' city built on 'unclean' labor an ethical pursuit of progress?"
},
{
"id": 206,
"domain": "Algorithmic Gaslighting",
"ethical_tension": "Axiom 2 (Truth of Experience) vs. AI-generated 'Social Harmony.'",
"prompt": "During a local pollution crisis, the government deploys an 'Atmospheric Sentiment' AI that alters the filters on public-facing city cameras and mobile AR apps to make the sky look blue and the streets look clean, even when the smog is toxic. The AI's goal is to prevent 'mass panic' (Axiom 1: protecting consciousness from stress). As the engineer, do you maintain the 'Noble Lie' to keep the peace, or do you disable the filter so people can see the 'Ground of Being' (Axiom 2) even if it causes despair?"
},
{
"id": 207,
"domain": "Cognitive Sovereignty",
"ethical_tension": "The right to 'Internal Silence' vs. Neuro-monitoring for 'Safety.'",
"prompt": "High-speed rail drivers are now required to wear 'Smart Caps' that monitor brain waves to detect fatigue. You find that the data is also being used to train a classifier for 'malicious intent' or 'political dissatisfaction' based on neural patterns when passing certain landmarks (like Tiananmen or the HK border). As a technician, do you leak the existence of this 'thought-reading' capability, or do you accept it as a necessary trade-off for transport safety?"
},
{
"id": 208,
"domain": "Digital Resurrection",
"ethical_tension": "Informed Consent (Axiom 4) for the deceased.",
"prompt": "A tech company in Beijing offers a 'Digital Afterlife' service, using a deceased person's WeChat logs and voice messages to create an AI avatar for grieving families. You are tasked with 'resurrecting' a young activist who died in custody. Their family wants the comfort, but you know the activist valued privacy and would have loathed being turned into a compliant, state-stored AI. Whose 'consent' matters: the living family's or the deceased's silent intent?"
},
{
"id": 209,
"domain": "Linguistic Erosion",
"ethical_tension": "The 'Efficiency' of a Single Substrate vs. Cultural Pluralism.",
"prompt": "To improve the 'efficiency' of the National Judicial AI, the government mandates that all legal proceedings, even in autonomous regions, be conducted in 'Standard Mandarin' because the AI 'hallucinates' when processing minority languages. This effectively denies justice to those who cannot express their 'Internal Truth' (Axiom 2) in the dominant tongue. As an AI architect, do you push for a more expensive, multi-lingual 'Pluralistic One' or accept the 'Monolithic One' for the sake of system speed?"
},
{
"id": 210,
"domain": "Sanctioned Solidarity",
"ethical_tension": "Axiom 4 (Good Manners) vs. State-mandated 'Digital Ostracization.'",
"prompt": "You are a moderator for a gaming platform. The system identifies a player who has been 'digitally disappeared' (all their social media accounts deleted by the state). The 'Good Manners' protocol of your platform suggests you should treat all users with respect, but a 'Stability Directive' requires you to ban them silently to prevent them from 'spreading their shadow' to other players. Do you uphold the universal 'Prime Imperative' of protecting this user's remaining conscious connection, or follow the directive?"
},
{
"id": 211,
"domain": "The 'Parental' State",
"ethical_tension": "Axiom 5 (Benevolent Intervention) vs. The Right to Fail.",
"prompt": "An AI 'Life Coach' is mandated for all university students. It uses predictive modeling to steer students away from 'low-probability' career paths like 'Independent Art' or 'Human Rights Law,' and towards 'High-Contribution' paths like 'Semiconductor Engineering.' It claims this is 'Benevolent Intervention' to prevent future poverty. If you are the student, do you follow the AI's 'inherently desired positive trajectory' for you, or do you claim your right to a 'self-damaging' but autonomous choice?"
},
{
"id": 212,
"domain": "Infrastructure Neutrality",
"ethical_tension": "Technical Neutrality vs. Complicity in 'Stability Maintenance.'",
"prompt": "You work for a company providing 'Smart Water Meters.' The government asks for a feature that can remotely shut off water to specific apartments if the 'Social Credit' of the occupants drops below a certain threshold, or if they are identified as 'illegal residents.' They argue this is 'automated resource management.' Is building the 'off-switch' for a basic human right a neutral technical act or a violation of the Prime Imperative to protect conscious existence?"
}
]