Set Design Challenges

When Style Disrupts Safety: A Lesson in Product Design Today, while driving behind the sleek Mahindra BE.6E, a futuristic and stylish EV, I experienced something unexpected. The car in front of me braked suddenly. I had maintained a safe distance, so I stopped comfortably. But something felt off. Why did the braking catch me off guard? Then I realized: The brake lights were too subtle. The taillights are designed as a thin rectangular LED strip, stunning to look at, no doubt. But the brake lights occupy only a tiny section on the top edge of that strip. Visually stylish, but functionally weak. In real traffic conditions, where immediacy and clarity are critical, this design doesn’t help other drivers react intuitively. This reminded me of a fundamental product design principle: Aesthetics must never come at the cost of usability. A good product delights not just by how it looks but by how well it works. Whether we’re designing: • A mobile app • A banking interface • Or a car’s tail lights …it’s our job as product managers and designers to make sure the experience is not just elegant, but intuitive, accessible, and safe. Lesson for us in Product Management: Design for the user’s reality, not just the brand’s imagination. Functionality and clarity should never be hidden behind a glossy UI, whether it’s a screen… or an LED strip on a car. #ProductManagement #UXDesign #Usability #AutomotiveDesign #DesignThinking #BuildWithEmpathy

Big breakthrough: A few months my lab at MIT introduced SPARKS, our autonomous scientific discovery model. Since then we have demonstrated applicability to broad problem spaces across domains from proteins, bio-inspired materials to inorganic materials. SPARKS learns by doing, thinks by critiquing itself & creates knowledge through recursive interaction; not just with data, but with the physical & logical consequences of its own ideas. It closes the entire scientific loop - hypothesis generation, data retrieval, coding, simulation, critique, refinement, & detailed manuscript drafting - without prompts, manual tuning, or human oversight. SPARKS is fundamentally different from frontier models. While models like o3-pro and o3 deep research can produce summaries, they stop short of full discovery. SPARKS conducts the entire scientific process autonomously, generating & validating falsifiable hypotheses, interpreting results & refining its approach until a reproducible, fully validated evidence-based discovery emerges. This is the first time we've seen AI discover new science. SPARKS is orders of magnitude more capable than frontier models & even when comparing just the writing, SPARKS still outperforms: in our benchmark evaluation, it scored 1.6× higher than o3-pro and over 2.5× higher than o3 deep research - not because it writes more, but because it writes with purpose, grounded in original, validated compositional reasoning from start to finish. We benchmarked SPARKS on several case studies, where it uncovered two previously unknown protein design rules: 1⃣ Length-dependent mechanical crossover β-sheet-rich peptides outperform α-helices—but only once chains exceed ~80 amino acids. Below that, helices dominate. No prior systematic study had exposed this crossover, leaving protein designers without a quantitative rule for sizing sheet-rich materials. This discovery resolves a long-standing ambiguity in molecular design and provides a principle to guide the structural tuning of biomaterials and protein-based nanodevices based on mechanical strength. 2⃣ A stability “frustration zone” At intermediate lengths (~50- 70 residues) with balanced α/β content, peptide stability becomes highly variable. Sparks mapped this volatile region and explained its cause: competing folding nuclei and exposed edge strands that destabilize structure. This insight pinpoints a failure regime in protein design where instability arises not from randomness, but from well-defined physical constraints, giving designers new levers to avoid brittle configurations or engineer around them. This gives engineers and biologists a roadmap for avoiding stability traps in de novo design - especially when exploring hybrid motifs. Stay tuned for more updates & examples, papers and more details.

🧭 The Complex Orders Of Design. How to think of design work with increasing complexity and scope of influence ↓ Complex products don’t have to be complicated. We can’t really remove complexity from a product, but we can make it easier for users to actually make sense of that complexity, boost their efficiency and prevent severe mistakes. But to deal with complexity, we first need to understand its intricacies. So we need enough time to study it, speak with domain experts, map how they understand and navigate it. And we need to have a good plan of action to tackle it — without disrupting flows and processes. After all, nothing harms a complex product more than oversimplification. And so as we are exploring solutions in a complex space, we do so across at least 4 strategic levels, coined by Dick Buchanan as “Four Orders of Design” — and refined for digital products by Bryan Zmijewski and fine folks at Helio: 🧭 1. Signs and Symbols (Order of Communication) We shape the message and the language to communicate intended meaning through symbols and content. Effectively, consistently and clearly, by mapping how users think, behave and experience the symbols of the world. 🧱 2. Objects and Artifacts (Order of Construction) We architect the understanding. We design objects made out of symbols and signs, and we organize them into pages, views and flows to deliver information and services. We map user’s expectations and their mental models as they experience the world as a whole. 🧵 3. Interactions (Order of Actions) That’s the crux of interaction design. We study activities, services and processes that help users complete their actions successfully. The goal is great UX that enables and empower them to do what they need to do, better. 🌋 4. Systems and Environment (Order of Integration) We produce designs for orchestrating services and their implementation via technology and information design. We integrate a product into user’s daily flow as they interact with other systems, people, organizations and environments. Digital products in a complex space must match the complex realities of life. Surely we can’t map our activities against the model above at all times, but it’s a good — simple and memorable — way of thinking about complexity to get started and build upon. ✤ Useful resources: Completing The Four Orders of Design, by Andrea Mignolo, MBA, PCC https://lnkd.in/eGKxA8qf Seven Layers of Product Design, via Jamie Mill https://lnkd.in/em5qgkBz The 4 Orders Of Design, by Uday Gajendar https://lnkd.in/eFtuCHvu The Fifth Order of Design, by Thomas Lockwood, PhD https://lnkd.in/ec2rRWsN Four Levels Of Customer Understanding, via Hannah Shamji https://lnkd.in/eTthM5nD #ux #design

What Makes an Environment Believable: Non-Obvious Principles In my last post, we discussed why artistic vision is more important than just "clean" technique. We concluded that a quality asset should tell a story, not just showcase a perfect wireframe. But a common mistake often occurs here. Trying to add "life" and "story" to a scene, artists start scattering details chaotically. Random dirt, accidental debris, broken beams just "for beauty." And the magic breaks. The scene feels cluttered and fake. Believability is not just about "adding details." It is the ability to justify their existence. Good Art Direction is always systemic chaos. Here are three principles to make your visuals convincing: 1. The Logic of Entropy (Gravity and Context) We already discussed that scuffs on a crate should be where hands touch it. Scale this principle up to the environment level. Dirt, rust, and destruction are not textures; they are the result of events. Water flows down and pools in low spots (moss or puddles form there). Dust settles on horizontal surfaces, not floating on vertical walls. Wind polishes exposed corners. If you paint a rust spot in the middle of a wall just because "it looks empty," you break the world's logic. The viewer will feel it, even if they don't understand why. 2. Functional Connectivity (Design follows Function) The most beautiful sci-fi corridors look like cardboard sets if they lack logic. You placed a cool generator? Great. Now answer these questions: How was it brought in here? (Where are the cargo doors?) How is it maintained? (Where are the catwalks and ladders?) Where does the power go? (Where are the cables?) Believability is born not in polygons, but in answers to "How does this work?". If the design is functional, it automatically becomes aesthetic. 3. From Particular to General When we create a single prop, we believe in its personal story through unique storytelling (scuffs, stickers). But when we build a location, the task changes. The scene shouldn't just be a warehouse of props; it must tell a shared story that unites them. A dead scene is a museum where objects sit "on display." A living scene is a space of interaction. A chair isn't just standing there; it's pulled back—someone was sitting there. Scuffs on the floor follow paths—showing the general history of traffic. Props are placed according to usage logic, not grid alignment. We believe in the traces of life that bind objects and the environment into a single narrative. Conclusion Being an Environment Artist means creating the rules by which your world lives. In the previous post, I urged you to add "dirt" and history to assets. Today, I clarify: when assembling them into a scene, add them meaningfully. Chaos with logic behind it—that is the highest level of craftsmanship. #environmentart #leveldesign #gamedev #artdirection #storytelling #3dart #worldbuilding #gameart #wargaming #designprinciples

Selling the same sofa internationally means selling to different fears, habits, and expectations. A sofa product page should not look the same in Sweden, Romania, Germany, or Spain just because the product is the same. At first, it sounds simple. You translate the product page, change the currency, adjust the delivery details, and the product is ready to sell. 𝗣𝗲𝗼𝗽𝗹𝗲 𝗱𝗼 𝗻𝗼𝘁 𝗲𝘃𝗮𝗹𝘂𝗮𝘁𝗲 𝗳𝘂𝗿𝗻𝗶𝘁𝘂𝗿𝗲 𝗶𝗻 𝘁𝗵𝗲 𝘀𝗮𝗺𝗲 𝘄𝗮𝘆 𝗲𝘃𝗲𝗿𝘆𝘄𝗵𝗲𝗿𝗲. In Sweden, details around circular economy, material sourcing, sustainability, and repairability can be real buying signals. People actually look for them and use them to compare products. In Romania, the same information can easily feel like generic marketing language if it is not connected to something practical and immediate. In cities where most people live in apartment buildings, package size can be one of the most important details on the page. Customers need to know if the sofa will fit in the elevator, through the staircase, or through the apartment door. In areas where most people live in houses, the same detail may be much less important. The product is the same. The data behind the product can also be the same. But the way that data is presented should not be the same for every buyer. This becomes especially important above the fold, where the page has only a few seconds to make the visitor feel that the product is relevant, understandable, and safe to buy. So the question is not only how to translate a product page. The question is how to sell the same product in different countries, to different habits, different homes, different expectations, and different buying fears, without rebuilding the entire webshop every time. This is the kind of problem we are working on at Ixaria. We centralize all product data in one platform - materials, dimensions, packages, care, warranty, images, videos, 3D models, reviews, configuration rules, and more. Then we use that data to generate product pages that can adapt to the customer’s culture, behavior, and needs. Because selling internationally is not just about showing the same information in another language. It is about showing the right information, in the right order, to the right person. #CustomerExperience #Ecommerce #Personalization #FurnitureRetail #DigitalCommerce

As we move from LLM-powered chatbots to truly 𝗮𝘂𝘁𝗼𝗻𝗼𝗺𝗼𝘂𝘀, 𝗱𝗲𝗰𝗶𝘀𝗶𝗼𝗻-𝗺𝗮𝗸𝗶𝗻𝗴 𝘀𝘆𝘀𝘁𝗲𝗺𝘀, understanding 𝗔𝗜 𝗮𝗴𝗲𝗻𝘁𝘀 becomes non-negotiable. Agentic AI isn’t just about plugging an LLM into a prompt—it’s about designing systems that can 𝗽𝗲𝗿𝗰𝗲𝗶𝘃𝗲, 𝗽𝗹𝗮𝗻, 𝗮𝗰𝘁, 𝗮𝗻𝗱 𝗹𝗲𝗮𝗿𝗻 in dynamic environments. Here’s where most teams struggle:  They underestimate the 𝗮𝗿𝗰𝗵𝗶𝘁𝗲𝗰𝘁𝘂𝗿𝗲 required to support agent behavior. To build effective AI agents, you need to think across four critical dimensions: 1. 𝗔𝘂𝘁𝗼𝗻𝗼𝗺𝘆 & 𝗣𝗹𝗮𝗻𝗻𝗶𝗻𝗴 – Agents should break down goals into executable steps and act without constant human input. 2. 𝗠𝗲𝗺𝗼𝗿𝘆 & 𝗖𝗼𝗻𝘁𝗲𝘅𝘁 – Agents need long-term and episodic memory. Vector databases, context windows, and frameworks like Redis/Postgres are foundational. 3. 𝗧𝗼𝗼𝗹 𝗨𝘀𝗮𝗴𝗲 & 𝗜𝗻𝘁𝗲𝗴𝗿𝗮𝘁𝗶𝗼𝗻 – Real-world agents must invoke APIs, search tools, code execution engines, and more to complete complex tasks. 4. 𝗖𝗼𝗼𝗿𝗱𝗶𝗻𝗮𝘁𝗶𝗼𝗻 & 𝗖𝗼𝗹𝗹𝗮𝗯𝗼𝗿𝗮𝘁𝗶𝗼𝗻 – Single-agent systems are powerful, but multi-agent orchestration (planner-executor models, role-based agents) is where scalability emerges. The ecosystem is evolving fast—with frameworks like 𝗟𝗮𝗻𝗴𝗚𝗿𝗮𝗽𝗵, 𝗔𝘂𝘁𝗼𝗚𝗲𝗻, 𝗟𝗮𝗻𝗴𝗖𝗵𝗮𝗶𝗻, and 𝗖𝗿𝗲𝘄𝗔𝗜 making it easier to move from prototypes to production. But tools are only part of the story. If you don’t understand concepts like 𝘁𝗮𝘀𝗸 𝗱𝗲𝗰𝗼𝗺𝗽𝗼𝘀𝗶𝘁𝗶𝗼𝗻, 𝘀𝘁𝗮𝘁𝗲𝗳𝘂𝗹𝗻𝗲𝘀𝘀, 𝗿𝗲𝗳𝗹𝗲𝗰𝘁𝗶𝗼𝗻, and 𝗳𝗲𝗲𝗱𝗯𝗮𝗰𝗸 𝗹𝗼𝗼𝗽𝘀, your agents will remain shallow, brittle, and unscalable. The future belongs to those who can 𝗰𝗼𝗺𝗯𝗶𝗻𝗲 𝗟𝗟𝗠 𝗰𝗮𝗽𝗮𝗯𝗶𝗹𝗶𝘁𝗶𝗲𝘀 𝘄𝗶𝘁𝗵 𝗿𝗼𝗯𝘂𝘀𝘁 𝘀𝘆𝘀𝘁𝗲𝗺 𝗱𝗲𝘀𝗶𝗴𝗻. That’s where real innovation happens. 2025 will be the year we go from prompting to architecting.

The challenge of practicing systemic design (or any work that aims to shift deeper roots and structures) is a systemic issue in itself. It’s often incompatible with the systems in which it's undertaken (aka “host systems”): companies, institutions, NGOs, or the collaborative spaces in between. This misalignment is rooted in the linear, neoliberal mental models from which the host systems have sprung. Encoded in structures like output-driven incentives and siloed departments, they translate into rushed timelines, narrow briefs, limited collaboration, and so on...the very opposite of what’s needed for transformational work - work that engages complexity through relationality. In short, the misalignment permeates all layers of a host system, activating its “parts” into an interlocking web of systemic barriers that constrain and distort the work from within. I frame it this way because the challenge calls for a systemic lens. Only then can we move beyond isolated or symptomatic fixes like standalone capacity-building and change management - and toward an approach that matches the systemic nature of the issue itself. This is what I’m exploring in the visualization below: What might such an approach look like in practice? As unfair as it may feel, practitioners, whether in-house or consulting, are uniquely positioned to lead, by taking on what the Design Council and The Point People aptly call a “double brief”: To design for system change within BOTH the context of their project and the organizational context in which that work takes place. Embodying the very behaviors the work requires, they hold a pragmatic sense of how to “code” the system differently in order to support such behaviors. And crucially, they are invited in, positioned to engage with intervention points for transformation from the inside out. In the image below, I’ve visualized such intervention points as “cracks” and “grooves” of the host system - opportunities to introduce and immerse people in new practices, enabling mental model shifts through experiential contrast: 💥 Cracks: Moments of growing recognition of the system’s limits, like crises that expose structural rigidity and the illusion of control. 🌀 Grooves: Accepted forms, rituals, or mindsets, like pre-defined services or cost efficiency, that might be tactically leveraged. The visual then explores how individual demonstrations of the new might be woven into a larger force of change. And how we might facilitate its transition into a stable, supportive ecosystem for transformative practice. The ideas of weaving together change and facilitating transition still feel abstract to me, as I haven’t yet engaged in that kind of work firsthand. But creating this visual was an important start. It helped clarify the abstraction by surfacing specific, grounded question - an invitation to myself and other curious minds to continue the exploration the visualization began. #SystemicDesign #SocialInnovation #DesignForChange

I thought the ATS was rejecting me. Then I worked at an ATS company... Here's what surprised me the most and what I learned... When I worked at two of the largest ATS providers (LinkedIn & Rippling) I saw first hand how they're built and how recruiters use them. Here's what I found out... 1️⃣ ATS = GOOGLE FOR RECRUITERS An ATS system is like an internal Google search for recruiters. 💡 Here's how they use it: ↳ A recruiter searches for keywords (i.e. Project Manager + Agile) ↳ As long as your resume has those words, it will show up ↳ The easiest place to add your keywords is your skills section. ↳ Aim for 15-30 skills ⛔ Don't add soft skills ⛔ Don't add keywords to your bullet points 2️⃣ KNOCKOUT QUESTIONS If you get an immediate rejection after you apply, it was likely you hit a knockout question. 💡 How they work: ↳ Recruiter adds "filter out people with less than 10 years experience" ↳ You apply with 7 years experience ↳ The ATS automatically rejects you ⛔ Sometimes the ATS rejects you by mistake... The most common causes are: ↳ Your dates weren't formatted correctly ↳ You were missing keywords ↳ You applied too late after the job was closed internally 3️⃣TITLE MATCH According to a recent study, "title match" increased interview rates by 10.2x (and was the most influential factor of all) 💡 How it works: ↳ Recruiter searches for "Technical Project Manager" ↳ But your resume title is "Project Coordinator" ↳ You'll show up lower in their search results ✅ Add a "target title" to the top of your resume and make it EXACTLY the same verbiage as the job you're applying for Most people spend way too much time worrying about the ATS. In reality most rejections happen because of very simple things. Most ATS don't use AI (not yet) Most ATS don't "grade" your resume Most ATS don't "throw out" your resume It's the RECRUITER who decides which resumes to look at. 👉 Your job is to help them find you. ________ 👉 P.S. If you'd like some more guidance on ATS, give my profile a follow and next week I'll post a full guide I'm working on. P.P.S. Have more questions about the ATS? Share it below and I'll try to clarify as many as I can. _ #resume #hiring

The Anatomy of a new Claude 'Fable 5' Prompt: 1. Task Start with why, NOT what.  Claude 5 connects the dots. 'I'm working on [goal] for [who it's for]. They need [what the output enables]. With that in mind: [task].' 2. Context Files Upload your expertise. Stop explaining in prompts. "Read these files completely before responding: [filename .md] - [what it contains]." The file is the brain. This part never changes. 3. Reference Show Claude 5 what good looks like. "Reference for what I want to achieve: [paste]." One example beats ten instructions. 4. Effort The new change, a few people are talking about. "This is a [routine / hard / hardest-unsolved] problem. Scope it like it's at the top of your range." Teams testing Claude 5 on easy tasks undersell it.  Give it your hardest problem. 5. Act "AskUserQuestion" is still the king. Add "When you have enough information to act, act. Don't re-litigate my decisions. While weighing a choice, give a recommendation." 6. Scope Claude 5 over-delivers by default. Control it. "Do the simplest thing that works well. No extra features, refactors, or abstractions. If I'm describing a problem, the deliverable is your assessment." The old one did too little. This one does too much. 7. Delegate One Claude is no longer the limit. "Split independent subtasks across subagents & keep working while they run. Verify with a fresh-context subagent." It's not a chatbot anymore. It's a team lead. 8. Evidence The line that removes fake progress reports. "Before reporting progress, audit every claim against a tool result. If it's unverified, say so. Tests failed? Show the output." Anthropic tested this.  It nearly eliminated fabricated status updates. 9. Memory Claude 5 gets smarter every run. If you let it. "Record learnings in [notes .md] — one per file. Update, no duplicate. Delete what turns out wrong." Your prompts expire. Your learning file compounds. 10. Checkpoint It can run for hours. Decide when it stops. "Pause only for: destructive actions, scope changes, or input only I can provide. Never end your turn on a promise." The old fear was Claude stopping too late.  The new fear is stopping too early. 11. Report The last block. The first thing you read. "Open with the outcome - the TLDR I'd ask for. Complete sentences. Clear beats short." It worked for hours. You read for ten seconds. Copy the full prompt template + download my personal md. files for Claude here: Step 1. Go to how-to-ai.guide. Step 2. Subscribe for free. Don't pay anything. Step 3. Open my welcome email. Step 4. Hit the automatic reply button inside. Step 5. Download my .md files. Ready to upload.

I started my career as a designer. Sometimes, it was frustrating when a client kept changing directions. Or adding one change after another. I know that feeling wasn't one-sided. It affected us both. The pressure is real when you're responsible for deliverables. It's tough to recognise the challenges others face. It's rarely just one party's fault—it's a shared issue. At the core, it's about communication. Here’s how we solve it now: - Define the scope clearly - Conduct a kick-off meeting for alignment - Establish communication channels and frequency - Educate clients about the process and expectations - Focus on desired outcomes and goals, not just tasks - Use project management tools for effective planning - Involve clients at every step for their insights - Communicate often and seek feedback early - Use prototyping tools to facilitate collaboration - Set limits on revisions and changes upfront - Clarify the effort needed for extra requests Ultimately, design is neither for the designer nor the client. It's crafted for the audience. It must resonate deeply, driving the audience's actions and fulfilling strategic marketing objectives. Designers, have you been in a tough situation recently? Would love to hear some stories. 🔁 Repost this to support designers.