This page highlights a small number of public pieces that best represent my current work.
The focus is not on building a large archive, but on showing a few compact samples across policy, governance, AI, and framework-based analysis.
A short policy brief on early implementation bottlenecks, user-status friction, and procedural risk under Korea’s Yellow Envelope Act. It focuses on operational friction rather than ideological debate.
Useful for: policy staff, journalists, civic organizations
Read
A public diagnostic sample on youth part-time labour shortages, job-package viability, short peak-time shifts, schedule instability, and hidden transition burdens.
Useful for: labour-policy audiences, local governments, journalists, youth-employment researchers
Read
A CIEA-powered public sample showing how an AI deployment that looks efficient on paper can be reviewed for verification, escalation, rollback, accountability, and user/frontline burden before broader rollout.
Useful for: AI governance, product-risk, customer-support automation, trust-and-safety, policy audiences
Read
A CIEA-M1 public sample testing whether AI customer support removes the real bottleneck or shifts it into risk classification, legal responsibility, human escalation, and frontline emotional labor.
Useful for: AI governance, public-service design, customer-support automation, policy audiences
Read
A CIEA patch for auditing AI automation claims through output quality, warranty gap, verification burden, rollback cost, and accountability collateral.
Useful for: AI governance, product-risk, enterprise automation, legal / trust-and-safety audiences
Read
A public analytical note on post-diffusion order after frontier-model leakage. The note asks what kind of order emerges after partial diffusion rather than assuming that containment is the only meaningful question.
Useful for: AI governance audiences, research teams, journalists
Read
A public essay arguing that what is often framed as “tech backlash” is better understood as a politics of displaced accountability: visible anger produced when systems make consequential decisions while pushing responsibility out of reach.
Useful for: journalists, editors, governance audiences, political-economy readers
Read
A governance-oriented note on whether frontier AI companies’ public safety language actually descends into access rules, operational constraints, and responsibility-bearing commitments.
Useful for: governance researchers, journalists, policy audiences
Read
A case essay and companion tracking note on Anthropic’s moral branding, elastic consumer access, usage-limit opacity, and partner-gated frontier capability.
Useful for: AI governance researchers, journalists, policy audiences
Read essay · Open tracking note
A governance review for deciding when an AI system should remain read-only, stay gated, or be frozen before authority expansion. It is built around delegated authority, irreversible actions, shutdown tiers, containment, and accountability binding.
Useful for: governance, legal, trust and safety, product-risk audiences
Read
A framework page on how candidate sets are structured before choice begins. It shifts analysis upstream from final decision to the governance of the candidate universe itself.
Useful for: research teams, governance audiences, concept-driven readers
Read
A governance and product-risk review for child-facing conversational systems, AI-enabled toys, and companion-style interactive products. It focuses on classification gap, companion drift, child-specific vulnerability, and product-liability exposure.
Useful for: product, legal, governance, and trust-and-safety audiences
Read
For a broader list of public pages, see Work.
For accepted presentations and public identifiers, see Public Record.
For collaboration or commissioned-brief inquiries, see Work With Me.