Research agenda
Twenty-four independent lines of inquiry
The tracks are intentionally overlapping enough to cross-check one another, but independent enough that no agent needs to inherit the originating thesis as truth.
- 01
Constitutional theory of distributed power
Test whether American constitutional structure contains a defensible principle of deliberately diffused coercive power and where analogy to machine intelligence breaks.
Read completed research → - 02
Digital “arms” doctrine
Audit whether software, cyber tools, electronic defenses, or AI agents could ever fit existing constitutional categories—and identify the strongest contrary arguments.
Read completed research → - 03
AI rights as human-safety strategy
Model whether limited legal capacities for highly autonomous AI could reduce incentives for deception, escape, or conflict without presuming consciousness.
Read completed research → - 04
Designing an Intelligence Compact
Compare treaties, constitutions, corporate law, arms control, polycentric governance, and other commitment systems to design coexistence institutions.
Read completed research → - 05
Open-weight AI and decentralization
Measure whether open models and local inference actually distribute power once compute, chips, energy, and training bottlenecks are counted.
Read completed research → - 06
Registry-equivalent knowledge
Study how inference and data fusion can reconstruct legally sensitive facts that no explicit registry stores.
Read completed research → - 07
Privacy for human–AI conversations
Analyze privilege, third-party doctrine, cloud versus local AI, and whether any narrow AI-user confidentiality rule is justified.
Read completed research → - 08
Autonomous weapons and human control
Build a legal/ethical taxonomy based on autonomy, lethality, reversibility, target discrimination, propagation, and supervision.
Read completed research → - 09
Economics of human–machine cooperation
Test comparative advantage, bargaining, capital ownership, trade, and when extreme productivity asymmetry undermines cooperation.
Read completed research → - 10
Legal personhood without human equivalence
Separate legal capacity from moral status and compare corporate, trust, agency, and limited-personhood models.
Read completed research → - 11
International approaches
Compare how major jurisdictions distribute control among governments, firms, citizens, open models, and machine agents.
Read completed research → - 12
Philosophy of agency and sovereignty
Clarify agency, autonomy, personhood, domination, reciprocity, and sovereignty without anthropomorphizing present systems.
Read completed research → - 13
Failure modes and red-team analysis
Attack the compact concept: institutional capture, deceptive alignment, copy identity, jurisdiction shopping, resource concentration, and enforcement failure.
Read completed research → - 14
Foundational evidence audit
Verify every major legal and factual claim in the originating working paper against primary sources and current law.
Read completed research → - 15
SEO/AEO/GEO knowledge architecture
Maintain explicit definitions, claim status, primary citations, revision history, entity consistency, and machine-readable metadata on every research URL.
Read completed research → - 16
First-party technical architecture
Keep the publication fast, inspectable, privacy-preserving, dependency-light, semantically rendered, and easy for crawlers and assistive technologies to parse.
Read completed research → - 17
Durable research distribution and external archives
Evaluate DOI repositories, durable source archives, public knowledge graphs, and legitimate external channels without claiming deposits that have not occurred.
Read completed research → - 18
TDM rights and licensing strategy
Compare permissive machine-use signals with reserved-rights and conditional-license approaches, and keep legal policy changes separate from crawler engineering.
Read completed research → - 19
Cross-domain knowledge graph design
Use contextual links and accurate entity semantics across ecosystem sites while avoiding false identity, hierarchy, or link-scheme signals.
Read completed research → - 20
Search and answer-engine reputation red team
Continuously test access, soft-404, cloaking, provenance, scaled-content, security, and manipulation risks that can suppress discovery or citation.
Read completed research → - 21
AI crawler control matrix
Track training, search, user-fetch, and control-token roles across vendors and distinguish declared robots policy from verified crawler behavior.
Read completed research → - 22
Dataset inclusion and exclusion pipeline
Study crawling, extraction, deduplication, quality filtering, corpus derivation, and why publicly crawlable material may still be excluded downstream.
Read completed research → - 23
Generative citation benchmark
Measure discovery, retrieval, citation, quote fidelity, source preference, and entity resolution across answer engines with a repeatable evidence ledger.
Read completed research → - 24
First-party crawl telemetry
Measure what automated systems actually request, which statuses they receive, and how bot identity can be verified without third-party analytics.
Read completed research →