Schema Design

In the era of AI, trust is the currency. We can no longer tell if images are AI-generated. Language models confidently hallucinate facts that don’t exist. And generated outputs are inherently inconsistent—the same prompt yields different results each time. Data visualization has grappled with these questions for a long time. How do you make something that’s inherently constructed feel trustworthy? How do you show complexity without overwhelming? How do you empower people to draw their own conclusions? In this piece, I explore how eight principles from data visualization can help us design AI-native experiences that people can actually trust. Read the article: https://lnkd.in/g3x4nSns #AI #DesignThinking #DataVisualization #Trust

Design is at an inflection point. In the era of AI, we’re shifting from crafting static interfaces to choreographing behaviors. We’re designing how humans and AI work together. Most AI interfaces today follow the familiar chatbot pattern: ask, respond, repeat. But this barely scratches the surface of what”s possible when we treat AI as a true collaborative partner. In this latest piece, I introduce seven interaction models that form the foundation of “Intelligent Interfaces”—systems designed for genuine human-AI collaboration: 🎨 Artifact Creation — Collaborative canvases where humans author and AI assists  ⚙️ Process Completion — Workflows where AI handles routine advancement, humans make critical decisions 🎯 Goal Planning — Outcome-focused systems that turn objectives into actionable plans  🗺️ Landscape Generation — Knowledge rendered as explorable terrain  🌐 Ecosystems Exploration — Modeling complex systems with real-time feedback  📡 Signal Monitoring — Event streams with intelligent classification  👤 Persona Interrogation — Embodied AI partners with consistent character and memory Each lens represents a different pattern of collaboration, optimized for specific types of work and thinking. Together, they become a toolkit for designing AI-native experiences. We’ve also created a FigJam template with all seven lenses that you can duplicate and use in your own projects: 👉 sche.ma/toolkit The future belongs to designers who can envision and prototype these collaborative experiences. What new patterns are you discovering in your work? Read the full article: https://lnkd.in/gswQ5YtU #AI #UX #IntelligentInterfaces #DesignThinking #HumanCenteredAI

Factory.ai had the most elegant interface I'd seen in months. Clean sessions, thoughtful context visualization, beautiful design. It also completely broke my workflow. Every click to manage sessions pulled me out of problem-solving mode. Every visual scan of context panels interrupted my thinking. The tooling got in the way of the work. This led me to collaborate with Sumin Chou from Schema Design on something I'm calling "The Smart UI Revolution." Our analysis reveals a fundamental shift happening in AI tool design: we're moving from graphical UI-first thinking to contextually intelligent interfaces that adapt to cognitive flow rather than forcing interface management. The research backs this up. Recent studies show: → 23% improvement in interface adaptation precision with context-aware systems → 31% better task completion when UIs auto-adjust to user context → 67% reduction in cognitive load through smart adaptation But here's what excites me most: You absolutely need pixel-perfect image controls for design work. The difference? Smart interfaces will surface those tools contextually, configured appropriately, exactly when needed. We're not eliminating interfaces. We're making them intelligent. Goodbye, Graphical UI. Hello, Smart UI. Full analysis with Sumin (including academic research) in the comments 👇 #AIDesign #UserInterface #SmartUI #HumanComputerInteraction #UXDesign

For builders in AI, design is the edge. Because AI is largely invisible, the way a product feels to use—how it fits into our lives—matters more than ever. Experience drives growth more than efficiency, speed, or having the best model. We're at the vanguard of something entirely new with AI. The old language of interfaces—buttons, menus, dashboards—isn’t enough to envision, let alone create, what’s emerging. We’re in uncharted territory where the only path forward is creativity. Dreaming up entirely new ways for people to work together with AI that didn’t exist before. It’s about being intentional about the world we want. Not fitting the world into the AI narrative. It isn’t about the technology alone—but how it can empower us to be better. More human. Whether managing agents or co-creating artifacts or navigating physical and digital worlds, we need to come up with the new idioms for how we experience AI. The only thing that is certain is that nothing will be the same. Let’s build.

“The future is like new terrain that doesn’t make sense... yet.” Heard a great presentation on Friday by Julian Bleecker of Near Future Laboratory at a workshop called “Designing the Future of Aging” at IDEO. His opening remarks helped me connect the dots on why design matters so much for AI. Julian talked about design fiction and making the extraordinary ordinary—taking radical possibilities and making them feel like natural extensions of our world. This may be what is currently missing in the space of generative AI. We’re overwhelming people with what the models are capable of, while we simultaneously miss helping them understand how AI fits into their actual lives. We’re focusing on the technology instead of showing them the future that the technology enables. We need to represent what the world might feel like, not just what it can do. And like Julian pointed out: within infinite worlds of possibility we need to find the connections to everyday life. In other words: design matters. This current AI inflection point we are living through is about application. Stop leading with what the model can do. There is a lot of hype out there and we are still trying to fit everything into a singular narrative. Instead, focus on where this technology can actually fit into our lives and how it helps advance our aspirations. Give people a sense of experiences that feel familiar—even inevitable. The people and organizations who figure this out—and human-centered design is key—will be the ones that truly transform how we live and work. Thanks to Emily Privot McNamara and Tom Bassett for the invitation to join this event! #DesignFiction #AIUX #FutureDesign #HumanCenteredDesign

Left inspired by yesterday’s panels at “Human-Centered Design in a Machine-Centered Future” co-organized by TEDAI San Francisco and IDEO. Below are some of my key takeaways. 🎯 Human Agency vs. Technological Determinism: There‘s a dangerous rhetoric of inevitability around AI, but it‘s simply not true—we can change the narrative. We‘re in danger of force-fitting our definitions of good art, culture, and society to match the AI narrative—instead of the other way around. Ovetta Sampson made an important observation: “Unless we are intentional about creating the future we want, we will repeat the inequities of the world we live in today.” 🖐️ Embodied vs. Disembodied Intelligence: Ed Newton-Rex brought up the creative cost of delegation. When we hand off our drafts to AI, we lose something essential in the process. Going a step further, the conversation shifted to how writing or drawing by hand stimulates thinking in ways that computers (not to mention AI copilots) cannot replicate. Our intelligence is embodied—and we’re abstracting it away without considering the cost. 💰 Scarcity vs. Abundance Mindsets: Alvin Wang Graylin (汪丛青) attempted to reframe the entire debate: instead of a scarcity mindset and fighting over limited resources, we need to shift to an abundance mindset. Instead of thinking about theft and leverage, we should think instead about the value of co-creation. This vision offers a different path forward where AI will enable everyone to join the creator class. ⚡ Productivity vs. Creativity: A long time ago, personal computers were supposed to make our lives easy and luxurious. We may be facing the same false promise with AI. The question isn’t whether AI will boost productivity—it’s whether those gains will create “another era of drudgery” or actually free us up to pursue activities that feed our curiosity, creativity, and sense of purpose. 🎨 Design is a Choice: There was a repeated emphasis on “Design Matters.” Design choices today will determine whether AI makes us addicted or empowered, stupider or smarter. Pat Pataranutaporn asked: “Are we engineering our own stupidity? We’re reading less and have lower attention spans—if we aren’t learning, what kind of society will we become?” David Webster argued the other side, that AI can actually help us learn faster than ever before. Tools like #NotebookLM let us get smart on any topic quickly, multiplying our human potential rather than diminishing it. The question that will define our future is: What kind of people will we become with AI? The answer lies not in the technology itself, but in how intentionally we design our relationship with it. How are you thinking about the relationship between humans and AI? #AIUX #FutureOfWork #HumanCenteredDesign

I’ve been experimenting with Figma Make since its release, and I’m genuinely impressed by its potential for generating rapid prototypes of data-driven experiences. I built an interactive 3D globe visualization mapping the 250 largest cities worldwide by population and GDP with just a few prompts. Make was able to dial in the visual styles remarkably close to what I was envisioning. Key observations:  ✅ Incredibly fast iteration on complex visualizations ✅ Ability to handle data generation, mapping, and visual styling ✅ Seamless recovery from issues via version rollback Current limitations: ⚠️ Performance degrades as codebase grows (recompilation takes progressively longer) ⚠️ Dataset size constraints (500+ cities caused compilation failures) ⚠️ Occasional unresolvable issues requiring rollbacks Despite these issues, this feels like a significant step forward especially for visual prototyping from where these tools were only months ago. What’s your experience been with vibecoding and are you seeing similar potential? #AIUX #IntelligentInterfaces #HumanAI #vibecoding #creativecoding