Intervention Atlas

Explore practical ideas for AI governance, collective epistemics, digital minds, revisability and other trajectory-shaping work. Each brief shows sources and uncertainty.

• Independent character audits
• A model-spec ambiguity register
• Honesty under pressure evaluations
• Public input into behavioral principles
• Boundaries of deference
• Pro-social behavior beyond agreeable language
• Circuit breakers for concentrated command
• Audits for hidden loyalties
• Plural procurement for critical AI
• Shared authorization for consequential compute
• Protected routes for safety concerns
• A public-interest AI threat model
• Explicitly revisable commitments
• Institutions for moral learning
• Funding moral public goods
• Plural value elicitation
• Accountable citizen deliberation
• A reversibility audit
• Benchmarks for epistemic reliability
• Auditable forecasting infrastructure
• Research on appropriate trust
• An evidence-provenance workbench
• An assistant for better disagreement
• Future ground-truth bottlenecks
• Indicators relevant to digital welfare
• A welfare review process
• Compare possible welfare safeguards
• A moral-status evidence map
• Governance for possible digital interests
• Report the limits of welfare measures
• Map AI–space governance decisions
• Contestable access to space infrastructure
• Protection of orbital commons
• Revisable space commitments
• Sharing benefits from space resources
• A space-governance trigger monitor

Interactive tools load with JavaScript. Original sources retain their language and attribution.