Interactive, visual explainers of hard ideas — built with AI agents and LLMs, shaped by how humans actually learn. Explanations you don't just read, but poke, drag, and replay until they click.
Explorable Explanations is where I build interactive, visual explainers of hard ideas — the kind of thing you don't just read, but poke, drag, and replay until the concept clicks into place.
A diagram shows you a system frozen in time. A simulation lets you reach in and break it, then watch it heal. I'm interested in that second mode — explanations that hand you the controls, where understanding comes from doing rather than being told.
Each piece is built with AI agents and LLMs, and grounded in how humans actually learn — how we build intuition, chunk complexity, and grasp an idea by acting on it. The models do the heavy lifting of turning a dense subject into something interactive; the design follows what we know about attention, memory, and the way understanding forms.
What does it take to feel an idea, not just be told it's true?
Each one takes a single hard idea and builds an interactive piece around it — something you can scroll through, drag, and replay until it makes sense in your hands.