Josh Clark and Veronika Kindred joined developer Damien Filiatrault on the Commit & Push podcast for a conversation that gets concrete about how Sentient Design actually works: the patterns, the architecture, and the design decisions behind intelligent interfaces.
Josh and Veronika walk through the principles of Sentient Design and then get hands-on with examples. They demo Google’s Genie world model where experiences are generated on the fly based on the user’s actions. They show Salesforce’s Generative Canvas, which assembles dashboards based on what you need right now. And they walk through how to make AI translate user intent into a working UI.
Damien brings a developer’s healthy skepticism throughout: What happens when the dashboard rearranges while you’re reading it? How do you debug a UI that’s different for every user? Josh and Veronika’s answers are grounded in real constraints.
The conversation covered a lot of ground:
- What the “sentient” in Sentient Design means: interfaces that are aware of context and intent
- The principles behind Sentient Design: contextually aware, collaborative, deferential, ambient
- Promoting productive humility in systems and productive skepticism in users
- The “bespoke UI” design pattern: AI assembling design-system components on demand
- The “NPC” design pattern: AI participating as another user in multi-user interfaces
- Defensive design: guardrails and graceful recovery when AI stumbles
- Why the happy path is over (and what replaces it)
- Veronika’s “helpful impatience” with traditional best practices
Quotable
“There is no happy path anymore.” — Veronika Kindred
“We can create genuinely new kinds of experiences when we weave intelligence into the interface.” — Josh Clark
“We want to delegate decisions to AI, not abdicate them.” — Josh Clark
“Part of the Sentient Design philosophy is the ethics around using [AI]: employing things like ‘defensive design’ to make sure that users aren’t being thrown to the wolves by integrating AI into these systems.” — Veronika Kindred
