How I work with AI

AI has changed how I approach almost every stage of design work. I use it to synthesise research, stress-test ideas, write and iterate on copy, and prototype faster. I've explored Claude Code for building small functional things independently, and Figma MCP for closing the gap between design intent and what ships. It's less about automating tasks and more about staying in flow and moving with more confidence.

But the design thinking still has to come first. Output without understanding is just a faster way to build the wrong thing. AI is good at generating, not at knowing what actually matters to a user or why a product decision is the right one. That judgment comes from experience, research, and genuine curiosity about the problem. The tools amplify good thinking, they don't replace it.

The tools are moving fast and what's useful today might look different in six months, so I try to stay curious, keep experimenting, and hold my workflow loosely enough to adapt.