That align with existing mental models, like "Quiet AI" (invisible, background assistance) and "Folder Instructions" (setting intent once for a folder to auto-organize files, fill forms, or notify you). Value comes from reducing friction and mistakes through context-aware integration, not from adding new apps to learn
NNG: Crafting AI Explanations for Every Role in Your Enterprise
An NN/g framework for AI explainability in enterprises: three roles need different explanations — AI consultants/governance leads need global, system-level views; builders need local, interactive explanations for debugging; domain experts need plain-language, workflow-tied explanations. No single explanation fits all — explainability is a design problem, not a technical afterthought
AI: Never mind the prompts, here’s the thinking
A studio rebuilt its design process around AI — sprints stayed 5 days, but output got deeper by building all states at once and generating documentation from the working prototype. The real danger is "thinking debt" — AI never documents the why — so the process starts with an experience brief before any AI tool opens
Case Study: Everything You’ve Ever Agreed To (and Never Read)
A UW student team designed "Termsly" — a browser extension that uses AI to summarize Terms & Conditions with mood-based ratings and plain-language breakdowns, plus a "Terms Wrapped" annual recap of your data footprint. Users care about privacy but Terms are too long and confusing; Termsly makes consent glanceable, customizable, and actionable
@uxdigest
