UX Design And Artificial Intelligence

🔮 AI Interaction Design Patterns (https://www.shapeof.ai), a fantastic (!) living catalog of emerging design patterns, heuristics, anti-patterns and real-life examples that shape the experience of AI — from identifiers and wayfinding to prompts, tuners and trust indicators. Incredible project by incredible Emily Campbell. 👏🏼 👏🏽 👏🏾 AI experience can go way beyond a text box. One of the most underrated yet impactful patterns for AI interfaces is the ability to tune AI experiences. This could show itself as a style lenses or temperature knobs — little tools to help users generate a more personalized output easier. E.g. Risky ↔ Risk-averse, Sad ↔ Happy, Concrete ↔ Abstract, Creative ↔ Precise. Instead of expecting large and highly detailed text prompts, we could slow people down when they prompt — e.g. with prompt constructors, prompt strength meters, presets or templates. Perhaps by defining an expected format, structure, personas, roles as checkboxes or chips — both for user input and AI responses (priming). Another much-needed feature is scoping. Users should be able to quickly scope their inquiry to a particular domain, level of expertise, sources or even a set of videos or PDFs. We need pre-screening of sources, and proactive alignment with users. These are features that would make output much more specific without having to write a long prompt. And: the AI output shouldn’t be bulky nor static. Users should be able to granularly iterate or revise little bits of it — e.g. by asking for sources of specific statements, or diverging from one view to another, or manipulating small parts of an image or a video. These refinements should happen not via text prompts, but contextually — acting on the relevant parts of AI outcome. We can go way beyond a text prompt. Better results come from combining good old-fashioned design patterns such as search, filtering and sorting with AI — to first find relevant and trustworthy sources, and then generate insights from them. That’s a great way to boost accuracy and make AI more relevant to more people. 💎 Design Patterns For AI Interfaces Prompt UX Patterns, by Sharang Sharma https://lnkd.in/eCytfAe9 Where should AI sit in your UI?, by Sharang Sharma https://lnkd.in/dyyMKuU9 AI UX Patterns, by Luke Bennis https://lnkd.in/dF9AZeKZ Design Patterns For Building Trust, by If https://lnkd.in/eEJngtVv AI Design Patterns Catalogue, by Maggie Appleton https://lnkd.in/ebAp9Sb8 --- 🚀 Fantastic AI Examples: Elicit (research tables): https://elicit.com Consensus (confidence levels): https://consensus.app/ Scispace (search + AI): https://scispace.com v7 Labs (AI auto-fill): https://v7labs.com/ Exa (semantic grid): https://exa.ai DeepL (translation): https://deepl.com NotebookLM (scoping): https://notebooklm.google/ [continues in comments] #ux #ai

AI is no longer just decorating rooms. It’s redesigning how we live. AI can now rethink rooms, floors, and entire layouts—turning bold ideas into build-ready designs. Would you do floor like that? The data behind the shift: • 30–50% faster design cycles using generative layout tools • 100+ layout permutations generated from a single brief • Up to 20–30% improvement in space utilization • 10–25% energy savings when airflow, lighting, and thermal paths are simulated early • 40% fewer late-stage design changes thanks to digital testing What’s fundamentally different? AI treats floor plans like software systems: Pedestrian movement is simulated before construction Natural light and ventilation are optimized virtually Furniture, walls, and utilities are stress-tested digitally Cost, carbon footprint, and materials are optimized in parallel This enables: Smaller homes that feel larger Offices designed around productivity and wellbeing Buildings that adapt over time instead of aging poorly The biggest myth? AI replaces architects and designers. Reality: AI handles complexity and permutations. Humans focus on vision, culture, emotion, and identity. The future of architecture isn’t just smart. It’s generative, data-driven, and human-centric. #AI #Architecture #Design via @Visual Spaces Lab #PropTech #GenerativeAI #FutureOfLiving #SmartBuildings #Innovation

If you’re leading AI initiatives, here is a strategic cheat sheet to move from "𝗰𝗼𝗼𝗹 𝗱𝗲𝗺𝗼" to 𝗲𝗻𝘁𝗲𝗿𝗽𝗿𝗶𝘀𝗲 𝘃𝗮𝗹𝘂𝗲. Think Risk, ROI, and Scalability. This strategy moves you from "𝘄𝗲 𝗵𝗮𝘃𝗲 𝗮 𝗺𝗼𝗱𝗲𝗹" to "𝘄𝗲 𝗵𝗮𝘃𝗲 𝗮 𝗯𝘂𝘀𝗶𝗻𝗲𝘀𝘀 𝗮𝘀𝘀𝗲𝘁." 𝟭. 𝗧𝗵𝗲 "𝗪𝗵𝘆" 𝗚𝗮𝘁𝗲 (𝗣𝗿𝗲-𝗣𝗼𝗖) • Don’t build just because you can. Define the Business Problem first • Success: Is the potential value > 10x the estimated cost? • Decision: If the problem can be solved with Regex or SQL, kill the AI project now. 𝟮. 𝗧𝗵𝗲 𝗣𝗿𝗼𝗼𝗳 𝗼𝗳 𝗖𝗼𝗻𝗰𝗲𝗽𝘁 (𝗣𝗼𝗖) • Goal: Prove feasibility, not scalability. • Timebox: 4–6 weeks max. • Team: 1-2 AI Engineers + 1 Domain Expert (Data Scientist alone is not enough). • Metric: Technical feasibility (e.g., "Can the model actually predict X with >80% accuracy on historical data?") 𝟯. 𝗧𝗵𝗲 "𝗠𝗩𝗣" 𝗧𝗿𝗮𝗻𝘀𝗶𝘁𝗶𝗼𝗻 (𝗧𝗵𝗲 𝗩𝗮𝗹𝗹𝗲𝘆 𝗼𝗳 𝗗𝗲𝗮𝘁𝗵) • Shift from "Notebook" to "System." • Infrastructure: Move off local GPUs to a dev cloud environment. Containerize. • Data Pipeline: Replace manual CSV dumps with automated data ingestion. • Decision: Does the model work on new, unseen data? If accuracy drops >10%, halt and investigate "Data Drift." 𝟰. 𝗥𝗶𝘀𝗸 & 𝗚𝗼𝘃𝗲𝗿𝗻𝗮𝗻𝗰𝗲 (𝗧𝗵𝗲 "𝗟𝗮𝘄𝘆𝗲𝗿" 𝗣𝗵𝗮𝘀𝗲) • Compliance is not an afterthought. • Guardrails: Implement checks to prevent hallucination or toxic output (e.g., NeMo Guardrails, Guidance). • Risk Decision: What is the cost of a wrong answer? If high (e.g., medical advice), keep a "Human-in-the-Loop." 𝟱. 𝗣𝗿𝗼𝗱𝘂𝗰𝘁𝗶𝗼𝗻 𝗔𝗿𝗰𝗵𝗶𝘁𝗲𝗰𝘁𝘂𝗿𝗲 • Scalability & Latency: Users won’t wait 10 seconds for a token. • Serving: Use optimized inference engines (vLLM, TGI, Triton) • Cost Control: Implement token limits and caching. "Pay-as-you-go" can bankrupt you overnight if an API loop goes rogue. 𝟲. 𝗘𝘃𝗮𝗹𝘂𝗮𝘁𝗶𝗼𝗻 • Automated Eval: Use "LLM-as-a-Judge" to score outputs against a golden dataset. • Feedback Loops: Build a mechanism for users to Thumbs Up/Down outcomes. Gold for fine-tuning later. 𝟳. 𝗢𝗽𝗲𝗿𝗮𝘁𝗶𝗼𝗻𝘀 (𝗟𝗟𝗠𝗢𝗽𝘀) • Day 2 is harder than Day 1. • Observability: Trace chains and monitor latency/cost per request (LangSmith, Arize). • Retraining: Models rot. Define when to retrain (e.g., "When accuracy drops below 85%" or "Monthly"). 𝗧𝗲𝗮𝗺 𝗘𝘃𝗼𝗹𝘂𝘁𝗶𝗼𝗻 • PoC Phase: AI Engineer + Subject Matter Expert. • MVP Phase: + Data Engineer + Backend Engineer. • Production Phase: + MLOps Engineer + Product Manager + Legal/Compliance. 𝗛𝗼𝘄 𝘁𝗼 𝗺𝗮𝗻𝗮𝗴𝗲 𝗔𝗜 𝗣𝗿𝗼𝗷𝗲𝗰𝘁𝘀 (𝗺𝘆 𝗮𝗱𝘃𝗶𝗰𝗲): → Treat AI as a Product, not a Research Project. → Fail fast: A failed PoC cost $10k; a failed Production rollout costs $1M+. → Cost Modeling: Estimate inference costs at peak scale before you write a line of production code. What decision gates do you use in your AI roadmap? Follow Priyanka for more cloud and AI tips and tools #ai #aiforbusiness #aileadership

How can designers use Claude Code? Not as a chatbot. As a production engine! Tommaso Nervegna recently published a practical guide to move from static mockups to working software without becoming traditional developers. At first glance, it sounds like “AI helps you code.” It’s not that simple. This isn’t about asking AI to generate snippets and pasting them somewhere. It’s about using Claude Code as an execution layer, where design intent becomes runnable output. What’s happening in this workflow: 🔸 Designers describe outcomes, not syntax 🔸 Claude generates structured project scaffolding 🔸 Iteration happens conversationally, with persistent context 🔸 Components evolve into functional UI, not just visual artifacts 🔸 The feedback loop lives inside the AI workflow, not in Jira tickets That’s a different paradigm. This isn’t “design handoff improved.” It’s closer to: design-as-executable-logic. When AI understands the structure, constraints, and system intent, documentation becomes dynamic. It becomes operational. Still early? Definitely. Still messy? In parts. But directionally… this is big. Because if designers can reliably move from concept → structured logic → functional interface with AI as a collaborator, the bottleneck shifts. Less translation. More orchestration. More systems thinking. We’re getting closer to a world where: Design is infrastructure. Prompts are architecture. And iteration cycles collapse dramatically. 🔗 Check the Practical Guide: https://lnkd.in/d_C7Nad6 Would you use Claude Code as part of your design workflow, or does that blur a boundary you still want to keep? 👇 #DesignSystems #designsystem #ClaudeCode #GenerativeAI #AIDesign #DesignEngineering #DesignOps #ProductDesign #UXStrategy #VibeCoding

True Humans + AI work and thinking means humans should participate in the AI's thinking processes. A very interesting new paper proposes a "Collaborative Workshop" approach to extended chain-of-thought processes such as deep research. They base their approach on three principles: Transparency: The agent’s reasoning, file system, and terminal outputs are fully visible in real-time. Symmetrical Control: Humans and AI have equal authority to modify the workspace. A human can edit a code file or a plan document just as easily as the agent can. Role Fluidity: The workflow can seamlessly shift between AI-led (autonomous) and human-led (assisted) modes. Beyond the specifics of the approach outlined in this paper, these principles are excellent starting points for all AI interface design. They do this by externalizing the agent's thinking into a visible "Plan-as-Document" markdown file (TODO.md). Users can hit "Pause," edit the TODO.md file to correct the agent's strategy, and hit "Resume." The agent then reads the updated plan and adjusts immediately. Despite being designed for collaboration, the system proves highly capable autonomously. ResearStudio achieved 74.09% on the GAIA benchmark, outperforming OpenAI’s DeepResearch (67.36%) and other state-of-the-art systems. The paper gives concrete examples of how human participants in the collaborative thinking workflow create better results. "It transforms the agent from an opaque, brittle tool into a resilient, trustworthy partner, providing the essential safeguard needed to deploy autonomous systems on complex, real-world problems." Full code available with the paper. Image created by Nano Banana Pro

Will Apps Need to Redesign Their Interfaces to Accommodate AI Agents? AI agents from OpenAI, Perplexity, and others can comfortably navigate textual and structured digital spaces but quickly hit barriers when faced with visually oriented tools like Gamma, Canva, or WordPress. These popular applications were designed specifically for human cognitive styles, relying heavily on visual intuition, recognition of subtle cues, and interactions guided by visual metaphors. As we can see from early tests, an AI agent accessing these tools via a browser faces hurdles. The reason: interfaces designed around human perception and intuition become ambiguous or even indecipherable to a purely logic-driven entity. This poses a nuanced design question: to effectively support AI agents, will software companies need to consider creating specialised, agent-oriented interfaces separate from the human-focused UX? The idea isn’t simply about creating more structured web pages. Rather, it suggests building parallel visual experiences explicitly designed around AI cognition, incorporating clear functional signposting, predictable interactions, and logical progressions that agents can reliably parse. The implications are notable: ➡️ Strategic Differentiation: Platforms offering agent-friendly interfaces might attract companies prioritising automation and seamless AI integration, creating new competitive landscapes. ➡️ UX Complexity: App developers will need to strike a balance. How much complexity can they add before negatively impacting the human experience? Can dual interfaces coexist without excessive overhead? ➡️ Productivity and Innovation: With optimised interfaces, agents could more effectively handle complex workflows, opening up new productivity gains beyond basic task automation. Reflections: 🤔 Will AI-friendly UX design become a new competitive advantage? 🤔 How feasible is it for companies to maintain dual-interface platforms for humans and AI agents? 🤔 Will the cognitive divide between human intuition and AI logic become a central consideration in the next era of software design? I'd be very interested in your thoughts. #AI #UX #ProductDesign #FrictionAdvantage

How do materials fail, and how can we design stronger, tougher, and more resilient ones? Published in #PNAS, our physics-aware AI model integrates advanced reasoning, rational thinking, and strategic planning capabilities models with the ability to write and execute code, perform atomistic simulations to solicit new physics data from “first principles”, and conduct visual analysis of graphed results and molecular mechanisms. By employing a multiagent strategy, these capabilities are combined into an intelligent system designed to solve complex scientific analysis and design tasks, as applied here to alloy design and discovery. This is significant because our model overcomes the limitations of traditional data-driven approaches by integrating diverse AI capabilities—reasoning, simulations, and multimodal analysis—into a collaborative system, enabling autonomous, adaptive, and efficient solutions to complex, multiobjective materials design problems that were previously slow, expert-dependent, and domain-specific. Wonderful work by my postdoc Alireza Ghafarollahi! Background: The design of new alloys is a multiscale problem that requires a holistic approach that involves retrieving relevant knowledge, applying advanced computational methods, conducting experimental validations, and analyzing the results, a process that is typically slow and reserved for human experts. Machine learning can help accelerate this process, for instance, through the use of deep surrogate models that connect structural and chemical features to material properties, or vice versa. However, existing data-driven models often target specific material objectives, offering limited flexibility to integrate out-of-domain knowledge and cannot adapt to new, unforeseen challenges. Our model overcomes these limitations by leveraging the distinct capabilities of multiple AI agents that collaborate autonomously within a dynamic environment to solve complex materials design tasks. The proposed physics-aware generative AI platform, AtomAgents, synergizes the intelligence of LLMs and the dynamic collaboration among AI agents with expertise in various domains, incl. knowledge retrieval, multimodal data integration, physics-based simulations, and comprehensive results analysis across modalities. The concerted effort of the multiagent system allows for addressing complex materials design problems, as demonstrated by examples that include autonomously designing metallic alloys with enhanced properties compared to their pure counterparts. We demonstrate accurate prediction of key characteristics across alloys and highlight the crucial role of solid solution alloying to steer the development of alloys. Paper: https://lnkd.in/enusweMf Code: https://lnkd.in/eWv2eKwS MIT Schwarzman College of Computing MIT Civil and Environmental Engineering MIT Department of Mechanical Engineering (MechE) MIT Industrial Liaison Program MIT School of Engineering

“Agentic CX” isn’t a thing; at least it’s not as a useful label. (I know I’m going to get some heat for this one.) What is real though is the fact that the minute you put AI into a product or a service interaction, you create a new “someone on the other side” problem to solve for. That’s why I think the role question matters. Classic UX designers know flows & journeys, language, and intent. Classic engineers know systems, failure modes, and scale. AI deliverables need both, plus a third muscle most orgs aren’t staffing yet: designing for uncertainty, recovery, and proof. Having spent most of my career in financial services, let me offer a simple example: A bank adds an AI assistant to handle “Where’s my wire?” or “Why was I charged this fee?” The model answers the customer’s query quickly, but the customer actually needs clarity, maybe some next steps, and a clean path to action. The experience is still poor. So how do we move this forward? My recommendation for what to do next is to create a new hybrid seat in your company with real authority, let’s call it an “AI interaction (or journey) lead” who owns end-to-end behavior across design, engineering, and ops, including fallbacks, escalation, and accountability. This person should probably come out of product or design teams. What do you want your customers to feel when your company’s AI is wrong, but still in the proverbial room? Do you think that’s a design or engineering role/ challenge? #customerexperience #ai #automation

A Director of UX at a SaaS company recently shared a painful calculation with me: Their team of 3 researchers spent 75% of their time on manual analysis. At an average salary of $150K, that's nearly $300K annually spent on analyzing data. But the bigger cost? Critical product decisions made without insights because "we can't wait for research." Most UX and product teams are trapped in a costly cycle of inefficiency: Conduct user interviews → Spend 30+ hours manually analyzing → Create a report → Make decisions based on gut feeling before the report is ready. After watching UX teams struggle with this for years, I've identified the core problem: research insights are treated as artifacts, not conversations. This is why we built AI Wizard into Looppanel - a conversational research companion that transforms how teams extract value from user research. Instead of static reports and manual analysis, AI Wizard allows anyone to simply ask: "What pain points did users mention about the onboarding process?" "Summarize the key recommendations users suggested for improving the checkout flow." "What were the main differences in how novice users versus power users approached this task?" You start by selecting from templates like Pain Points, Recommendations, or Summary. AI Wizard instantly analyzes your project data and engages in a natural conversation - complete with follow-up questions to dig deeper into specific areas. The way I see it, AI Wizard helps solve 3 critical problems: 1. The speed-to-decision problem Waiting weeks for analysis means missing decision windows. AI Wizard delivers TLDR overviews in seconds, not days. 2. The iteration problem No more spending time on data again because of a follow-up question. Answer unexpected stakeholder questions on the spot instead of scheduling another week of analysis 3. The tailored communication problem Automatically format the same insights for different audiences: executives get metrics, designers get details, all without rebuilding presentations. With AI Wizard, your team can: → Start conversations with templates like Pain Points, Recommendations, or Summary → Ask follow-up questions to dig deeper → Get insights from across your entire research repository in seconds → Democratize access to insights throughout your organization Will your team be leading this transformation or catching up to it? If you want to make the shift, sign up for a personalized demo here: https://bit.ly/42PEOlX

Our designer Rishikesh spends an hour every day watching users get stuck. Every morning at 10 AM, he watches session recordings on 4x speed, looking for the moment users pause too long, click the same button twice, or try the same action again. Here’s a tiny example from last week. Users wanted to download their finished deck. But instead of clicking the download button, many of them were typing "download" into the agent. That is a product failure. The user should not have to ask the agent to do something the interface already supports. So Rishi made a small change: when users typed keywords like "download," we automatically surfaced the download button right where they needed it. The result: agent and support requests about downloading dropped to zero. Last month, we caught 14 patterns like this. Most were invisible in our analytics dashboard. This is why we watch users get stuck. Not because we enjoy watching people struggle. Because this is where the roadmap often hides. A support ticket tells you what a user could articulate. A behavior trace tells you what they actually experienced. The best product insights are often not in what customers say they want. They are in what customers keep trying to do despite your product fighting them. Founders/product teams: what is one thing your users keep trying to do that your product keeps fighting them on?