I am interested in systems that help people work with intelligence more deliberately. These are themes I want to study, test, and refine.
A WORKING MAP — 2026Research begins with better questions. This page records directions of inquiry, not completed studies or established results.
How can AI help extend human judgment without obscuring the decisions people need to make? I am drawn to workflows that expose assumptions, invite correction, and make room for different levels of agency.
OPEN QUESTION / WHERE SHOULD CONTROL LIVE?What design patterns help intelligent systems remain legible as context, memory, tools, and evaluation are combined? I want to explore architectures that can evolve without hiding how they work.
OPEN QUESTION / WHAT MAKES A SYSTEM UNDERSTANDABLE?How can an interface reveal the state and structure of a complex process at the right moment? My interest is in visual explanations that make systems easier to inspect, question, and steer.
OPEN QUESTION / WHAT SHOULD BECOME VISIBLE?