Thinking

A curated canon, not a feed. The pieces I'd point you to on the behavioral layer of AI, how systems act, earn trust, and recover when they're wrong. Each links out to where it lives.

Essays

Ranked · 10
Jun 2026Pillar · behavioral layer

AI Products Have a Behavioral Layer. Almost No One Owns It.

The layer that decides how AI signals confidence, escalates, and recovers is the new UX, and almost no one owns it.

Aug 2025Foundations

The Hidden Logic Layer: Where AI Design Actually Happens

The real design work moved below the interface, into the logic that governs how a system behaves. This names where it went.

Aug 2025AI leadership

The Biggest AI Mistake Design Leaders Are Making Right Now

Plus-AI versus AI-plus: bolting intelligence onto a product versus architecting around it, and why only one compounds.

Jun 2026Agentic interfaces

The Agentic Interface Has Two Faces. We're Building the Easy One.

Every agentic product serves humans and machines at once. The industry is building only the easy half.

Sep 2025Trust

How to Design Trust Into AI Systems (It's Not What You Think)

Trust isn't a feeling you hope for. It's engineered across five vectors: clarity, control, consistency, disclosure, repair.

Apr 2025Trust · research

Designing for Trust: Why Explainability Alone Falls Short in AI

The research says explainability barely moves trust. Consistency and repair are what earn it.

Oct 2025Behavioral contracts

Neuro-Symbolic AI: Why Behavioral Contracts Work When Prompting Fails

Why prompting alone can't make an agent behave, and what a testable behavioral contract does instead.

May 2026AI-native ops

The Self-Maintaining Framework: What AI-Native Operations Actually Look Like

What AI-native operations actually look like: human judgment authored once, then faithfully propagated by the system.

May 2025Agent experience

Agent UX (AUX): Designing for the Next User

Agents are users now. Designing for them is the next accessibility.

Feb 2026Policy & design

When Regulators Design Better Than We Do

A Chinese AI law did what most design principles can't: specify behavior precisely enough to enforce.

Full archive at blog.goldfoot.com →