Adres
Playground/ docs
AIFoundation

Introduction

Why AI experiences need explicit design guidelines — transparency, capability, and resilience.

AI Experience

Interfaces people
can trust.

AI chat is now a primary product touchpoint. Getting it wrong means confusion, broken trust, and users abandoning the interface. This section documents the guidelines, components, and patterns for building AI experiences that are honest, capable, and resilient.

TransparentCapableResilient

01 — Transparent

Never pretend to be human

AI must identify itself clearly from the first message. Disclose limitations proactively — not after users are already frustrated. Trust built on false impressions collapses catastrophically.

02 — Capable

Show what you can do

Users shouldn't have to guess what to type. Opening messages and suggested prompts should communicate scope and capability instantly — turning uncertainty into confident interaction.

03 — Resilient

Fail with grace

Every AI will fail. Good design means failing honestly, quickly recovering, and always providing a path to a human. An AI that admits uncertainty is more trustworthy than one that guesses.