Three arguments
I'm prepared to defend.

Clearview AI and the Limits of Regulatory Governance: Why Facial Recognition Markets Fail

Claim: facial-recognition markets can't be fixed by the regulation we have — only by regulation aimed at a different layer.

The essay applies three analytic frameworks to argue that facial recognition markets are structurally ungovernable through existing EU mechanisms: Akerlof's information asymmetry (buyers and regulators cannot evaluate what they are purchasing), Lukes' three-dimensional power model (companies shape not just decisions but the regulatory agenda itself), and Bradford's Brussels Effect (EU standards propagate globally, but surveillance technology companies route around them through jurisdictional arbitrage).

The core claim: the market failure Clearview AI represents is not correctable from within existing frameworks — it requires a structural intervention at the data-sourcing layer, not the deployment layer where current regulation focuses.

US–China Supply Chain Bifurcation: Validating the Creative Insecurity Thesis

Claim: sanctions on technology supply chains behave like industrial policy for the country being sanctioned.

The "creative insecurity" hypothesis holds that geopolitical pressure on technology supply chains functions as de facto industrial policy — forcing domestic innovation rather than merely restricting trade. This preprint tests that claim through Huawei's component localization data.

Finding: the implementation gap between stated US export-control goals (limiting Chinese capability) and actual outcomes (accelerating Chinese supply chain independence) is structurally predictable once you treat the restrictions as industrial incentives rather than capability denials.

Logistic Regression Modeling of G20 Trade Retaliation Patterns

Claim: trade retaliation isn't random — the escalation sequence has structure you can model.

A quantitative investigation into what predicts bilateral trade retaliation between G20 members, using logistic regression on historical tariff escalation sequences. The work provides the methodological counterpart to the qualitative arguments in the EE and preprint — demonstrating range across the quantitative/qualitative divide rather than treating them as separate domains.

the least glamorous exhibit, and the one that taught me the most about being wrong