My work sits between AI systems, political strategy, and the institutional questions that emerge when technical power moves faster than governance.

I am especially drawn to how competition shapes behavior: where coordination breaks down, how AI races distort incentives, and why those failures often fall hardest on countries with less bargaining power.

On the technical side, I work on evaluation and interpretability, trying to understand how models reason, where they fail, and what more rigorous assessment might look like in practice.