Bookshelf
Resources
Books, essays, and references I keep recommending — on design leadership, agentic AI, and the ethics of building things that act for people. Each one links out.
The UX of AI
A deeply pragmatic collection of design patterns that moves beyond the typical horizontal text box to tackle real-world constraints like model latency, error states, and confidence scores.
Thinking in Systems
I recommend this book to designers struggling with systems thinking. This is the mental model for how to build scalable abstractions and perhaps an explanation for why your design had the opposite effect from what you intended.
Thinking, Fast and Slow
The two-system model of the mind is the operating manual behind every default, nudge, and confirmation dialog you will ever design.
tldraw SDK
We’ve used tldraw SDK as the basis for our AI prototypes at Mural. It’s an infinite canvas out of the box, so we skip months of undifferentiated canvas work and ship prototypes that realistically test our new ideas.
Understanding Comics
The best book on visual communication, sequence, and closure, disguised as a comic. It rewires how you think about progressive disclosure and teaches you how to storyboard effectively.
We Don’t Sell Saddles Here
Stewart’s legendary Slack memo is a masterful lesson in framing software not by its mechanical features, but by the profound change in human capability it enables. I shared this with our executive team recently. Technology is shifting beneath our feet, and this piece serves as an inspiring reminder to stop designing mere software utilities and start designing the path to entirely new ways of working together.
When We Build
Steve Jobs’s idea of the computer is a “bicycle for the mind.” Wilson captures the blend of design, architecture, and technology on a designer’s desk: How the tools we build end up shaping our environment and ourselves.
Why Chatbots Are Not the Future
The sharpest critique of the blank text box as an AI interface, with real ideas for giving models affordances and handles. Core reading for agent UX.
You Look Like a Thing and I Love You
The funniest, most real primer on how AI actually fails. Hand it to any stakeholder who still thinks the model is magic.