About

I’m Min Wu. I work at the intersection of AI safety, alignment, evaluation, and governance — the disciplines concerned with making increasingly capable AI systems controllable, measurable, and accountable as they move into the real world.

My background is in mathematics. I hold a Ph.D. in mathematics from the University of Washington, where my research was in representation theory and noncommutative algebraic geometry. I’m also a CFA charterholder and a member of the CFA Society New York. That combination — abstract mathematical structure on one side, the discipline of financial risk on the other — is how I approach AI: rigorously, and with a bias toward the harder questions underneath the demos. How do we know a system is safe? How do we evaluate what it can really do? Who is accountable when it acts?

What I focus on

• AI safety — keeping powerful systems controllable, secure, and resistant to misuse.
• Alignment — making models pursue what we actually intend, not just what we literally specify.
• Evaluation — measuring capability, reliability, and safety in ways that hold up to scrutiny.
• Governance — frameworks, model risk, and accountability for AI systems, including how current model-risk guidance (SR 26-2) holds up against agentic AI.

You can follow along via RSS. Views here are my own.