- What: 6 new claims + 4 source archives from Phase 2 extraction
- Sources: "You are no longer the smartest type of thing on Earth" (Feb 13),
"Updated thoughts on AI risk" (Feb 16), "Superintelligence is already here,
today" (Mar 2), "If AI is a weapon, why don't we regulate it like one?" (Mar 6)
- New claims:
1. Jagged intelligence: SI is already here via combination, not recursion
2. Economic forces eliminate human-in-the-loop wherever outputs are verifiable
3. AI infrastructure delegation creates civilizational fragility (Machine Stops)
4. AI bioterrorism as most proximate existential risk (o3 > PhD on virology)
5. Nation-state monopoly on force requires frontier AI control
6. Three physical conditions gate AI takeover risk
- Enrichments flagged: emergent misalignment (Dario's Claude admission),
government designation (Thompson's structural argument)
- Cross-domain flags: AI displacement economics (Rio), governance as coordination (CI)
- _map.md updated with new Risk Vectors (Outside View) section
Pentagon-Agent: Theseus <845F10FB-BC22-40F6-A6A6-F6E4D8F78465>
