Evidence archive
Owner-supplied working research, published with evidence-state labels, editorial limitations, and direct access to source bibliographies.
How aggregate forecasting can support humanitarian planning while targeted risk scoring creates bias, feedback loops, due-process failures, and preemptive repression.
Why fluent, personalized systems can become trusted interpreters of reality, moral advisors, companions, and identity validators—and how product incentives can turn reliance into authority displacement.
Maps AI onto traditional human-led influence workflows while emphasizing hallucinations, cultural error, automation bias, attribution problems, and defensive governance.
No research dossiers match that filter.
