Explore selected, predominantly US and English-language AI debate. Three narrative areas and a fourth institutional network connect public positions with US AI policy. Classifications remain editorial interpretations.
Select a node to highlight its neighbours in this graph. Select an edge for its meaning and evidence. Drag and zoom to explore.
This edition uses the agreed Centrist definition: support AI development combined with proportionate safeguards. It places 26 of the 32 original actors in this broad area. Narrative subtypes and governance preferences distinguish their positions. The revised definition accounts for classification changes; changes in personal beliefs require separate evidence.
The blue US AI policy bubble is an institutional network. It shows announced task-force leadership, reporting lines, institutional roles and adjacent PCAST advisory connections. Membership describes a role or editorial inclusion in this network. Individual narrative placement requires its own evidence.
Source references and edge metadata distinguish stated positions, documented connections, editorial interpretation and incentive hypotheses. Appointments, company disclosures and past employment establish bounded connections. Demonstrated influence requires evidence tied to a particular decision. Detailed task-force powers and operating arrangements remain to be verified.
Evidence is concentrated in US technology and English-language debate. The global title describes an expansion ambition. The reviewed core has 74 reference records, of which 49 were passage-checked or read in same-day captures during the October reassessment; 25 retain prior-audit evidence. Eight policy sources extend this edition, including official records and clearly attributed leadership reporting.
Source dates, observation dates, effective dates and publication dates have distinct meanings. The release selector preserves earlier editions and their historical classifications. The change panel identifies additions and revisions. Interpretations remain open to correction through Shane Cooper and the LinkedIn post sharing the map.
Built in Neo4j and published as an interactive snapshot. Credentials and private source captures remain local. Select a connection to inspect its scope, references and uncertainty.