Panel
summary
Panel reports that an AI-aided language product called Finnish companion is built from scraped Wikipedia and Wiktionary data and serialized into a fast binary for on-device use, and that it already generates drills, audio loops and multiple-choice items. Car UI work targets touch, contrast and distraction rules and keeps features disabled while driving. Teams use Figma and a shared design system but struggle with component naming mismatches, design tokens naming causes friction, backward compatibility and modest analytics for measuring success.
AI-written summary · about
chapters (20)
- 0:00Introduction
- 1:30AI in design systems
- 4:20Automotive UI standards
- 6:10Finnish companion tools
- 7:35Design Sprints struggles
- 9:40Translation models naming
- 11:45Figma workflows
- 12:50AI for language learning
- 14:35Finnish companion data
- 16:00Interaction design tools
- 17:55Building Finnish companion
- 19:00Measuring learning methods
- 20:50Design for diverse learners
- 22:10User testing automotive UIs
- 22:55Sharing design systems
- 23:35Legacy devices design
- 25:10Legacy device considerations
- 26:10Safe programming languages
- 27:15AI in medicine
- 28:20LLM limitations in diagnostics
related talks
27mMisplAIced: Good UI Needs Good UX
43mPanel Discussion: Unraveling AI & Design
18mAI for UI
31mAI & UX: Designing the Future of Human Interaction
17mAI-Driven Frontend Development: Beyond Traditional
56mEnabling teams with AI
34mAutomating Full-Screen TalkBack for a Government App with Designer Collaboration and AI Agents — a Practical Experience
48mMastering the AI Shift: Adaptive Strategies for Software Evolution