Smashing Magazine
Day 45 after signup
AI Design & Engineering 2027
Why it works
- Open with a timely theme and a brief editorial introduction to give the issue a clear point of view.
- Organize recommendations into numbered sections so readers can scan a broad mix of articles and resources.
- Place event and book promotions alongside relevant editorial content to keep the issue useful as well as commercial.
How it is built
- Friendly greeting
- Conference and workshop promotions
- Editorial topic introduction
- Numbered article roundup
Email copy
AI-ready design system, design engineering skills, model context experience, agent experiences, AI UX Playground. Issue #584 • Nov 10, 2026 • View in the browser Aloha Smashing Friends, How do product teams work these days? Where do teams fail, and where do they succeed — especially when they try to integrate AI meaningfully in their work? Meet 🔮 AI Conf 2027, a friendly, practical, pragmatic half-day conference — on AI-ready design systems, AI workflows, design patterns for AI and the state of AI for 2027. With Brad Frost, Christine Vallaure, TJ Pitre and yours truly. Dec 1. Jump to the details. Meet AI 2027 Conference: a practical conference on state of AI 2027 and how to make the best out of it. Get your ticket. Until then, we’ve organized a good number of smashing online workshops for you in the following weeks to come: Design Patterns For AI Interfaces + self-paced course UX with Vitaly Friedman. Nov 16–24 AI-Ready Design Systems For Designers UX with Christine Vallaure. Nov 30–Dec 4 Designing Websites That Convert UX with Paul Boag. Dec 7 Designing For Complex UI Masterclass UX with Vitaly Friedman. December 2–16 UX Strategy Masterclass UX with Vitaly Friedman. December 11–15 Jump to all workshops → In this newsletter, we also look at AI design and engineering in 2027 and design workflows teams put in practice when integrating AI into their existing workflows. Let’s dive right in! — Vitaly 1. AI UX Playground How to design for models that stream, act autonomously, or handle tools? AI UX Playground features more than 150 AI interface patterns, demos, product teardowns, and guides to help you master the challenge. The collection also includes handpicked copy-paste prompts and agent skills you can use right away. 2. Agent Experience Optimization AI agents are changing the game: Instead of solely optimizing our sites and digital services for human visitors, we need to optimize for agents, too, to ensure they can discover and reference our content. Agent Experiences has got your back as you’re getting started with Agent Experience Optimization, with guides and resources and an Agent Analyzer to check how usable your site is for agents. 3. Stageshow Even good AI models still have difficulties producing a polished faux UI. To change that, Benji Taylor created Stageshow, a skill that gives your agent a workflowit can follow to achieve better results. The agent reads the skill, starts from the closest example, builds the demo, then films it and fixes what it finds. The result is an ordinary React component you can use and change like the rest of your code. 4. The Problem With The Sparkle Button The sparkle button seems to be almost everywhere these days — in spreadsheets, design tools, mail clients, team chats, video calls, and even banking apps. John Sabutis explores the problem with the omnipresent sparkle button, what the sparkle is trying to hide, and how we can do better. Do you have a meowment? 20 Years and Counting: Thanks For Being Smashing! Did you know that SmashingMag has been around for 20 years? Yep, time flies! If Smashing has helped you learn something, solve a problem, or see things from a new perspective, please do consider making a one-time contribution. Your support truly helps us keep practical web knowledge open and accessible to everyone. 🧡 5. AI-Ready Design System Developers can fill in the gaps when turning a Figma file into code. AI can’t do this reliably — it improvises when the instructions are unclear, and the result is often a component that looks finished on the surface but is partly invented. Christine Vallaure wrote a practical guide on how to start your design system in code instead and generate the Figma library from it with MCP to circumvent the problem. 6. Measuring AI-Readiness How AI-ready is our design system? To help design system teams answer the question, Christoph Hellmuth created open-design-system-bench. The benchmark runs real coding agents against