I Just Vacationed in One of My Favorite Design Systems
What a 1959 Palm Springs tract house taught me about design systems, Frank Lloyd Wright, and building something that still works after its author leaves the room.
More articles
My Second Brain Can Fix a Mistake. But It Still Needs Me to Notice One.
Across products, conversations, and trusted models, I want the context that matters to travel with me. Building that continuity has taught me how much judgment lives inside the act of remembering.
I spend most of my day inside Anthropic or OpenAI products. I move between Claude Code, Cowork, Codex, web,
The Agentic Interface Has Two Faces. We're Building the Easy One.
a16z says the interface is being rebuilt for agents. They're half right. The harder half is how agents behave with people, and almost no one owns it.
AI Products Have a Behavioral Layer. Almost No One Owns It.
Every AI product has a hidden layer that governs how it behaves. How it expresses confidence. When it escalates to a human. What it does when it is wrong. Whether it earns trust on the second interaction or loses it.
That layer is the product now. And in most companies, no one owns it.
The Self-Maintaining Framework: What AI-Native Operations Actually Look Like
Every morning at 5 AM Pacific, three agents of mine have already drafted their proposed updates to my framework. A fourth is finishing its work on the public website. By the time I sit down with coffee, the day's pull requests are queued for review.
When Regulators Design Better Than We Do: What China's AI Anthropomorphism Law Teaches Product Designers
I've spent years arguing that AI design happens at the behavioral layer, not the interface layer. That the real design decisions aren't about pixels and spacing but about how systems reason, communicate uncertainty, and earn trust. I wrote an entire book about it.
So when China
Neuro-Symbolic AI: Why Behavioral Contracts Work When Prompting Fails
Why prompting fails and how behavioral contracts combine neural intelligence with symbolic rules to create trustworthy, scalable AI systems.
The ChatGPT Data That Challenges Everything We Know About AI Design
The first comprehensive study of ChatGPT usage analyzed 2.5 billion daily messages and revealed a fundamental mismatch: 49% of interactions involve collaborative thinking, yet most AI design optimizes for task completion. We're building for efficiency when users seek understanding.
How to Design Trust Into AI Systems (It’s Not What You Think)
Trust isn't a feeling users have about AI—it's designed infrastructure. Learn the five vectors of AI trust (Clarity, Control, Consistency, Disclosure, Repair) that enable systematic human-AI collaboration at scale.
How to Lead Design in the AI Era: My New Book is Here
Stop fearing AI displacement. Learn how design leaders can become essential to organizational intelligence with the AI+ framework for strategic relevance.