<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0" xmlns:cc="http://cyber.law.harvard.edu/rss/creativeCommonsRssModule.html">
    <channel>
        <title><![CDATA[Stories by Dannie Sánchez on Medium]]></title>
        <description><![CDATA[Stories by Dannie Sánchez on Medium]]></description>
        <link>https://medium.com/@danniesnchez?source=rss-8f14711e9d01------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/0*XUvbFk5iscEjqsfZ.jpg</url>
            <title>Stories by Dannie Sánchez on Medium</title>
            <link>https://medium.com/@danniesnchez?source=rss-8f14711e9d01------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Sun, 22 Nov 2026 16:23:38 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@danniesnchez/feed" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[I Was Wrong. AI Can Design — and It’s Amazing.]]></title>
            <link>https://medium.com/@danniesnchez/i-was-wrong-ai-can-design-and-its-amazing-1f97f60677b0?source=rss-8f14711e9d01------2</link>
            <guid isPermaLink="false">https://medium.com/p/1f97f60677b0</guid>
            <dc:creator><![CDATA[Dannie Sánchez]]></dc:creator>
            <pubDate>Sun, 09 Nov 2025 22:42:36 GMT</pubDate>
            <atom:updated>2026-04-17T06:43:41.050Z</atom:updated>
            <content:encoded><![CDATA[<h3>What I Got Wrong About AI and Design</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*icZxzs6gU2Ek5V_m" /><figcaption>Photo by <a href="https://unsplash.com/@rtp_atw?utm_source=medium&amp;utm_medium=referral">Ratapan Anantawat</a> on <a href="https://unsplash.com/?utm_source=medium&amp;utm_medium=referral">Unsplash</a></figcaption></figure><p>Try this: open any AI tool and type “Design a great onboarding flow that gets users to see the value the product provides.” What comes back will look like a design. Screens, labels, a plausible sequence. It’ll be confident and reasonably coherent.</p><p>It’ll also be useless. Because it doesn’t know your users, your product’s actual value, your business constraints, or the specific moment where people tend to drop off. It’s a pattern that resembles something that worked somewhere else — reassembled for no one in particular.</p><p>I had to learn this the hard way. When I started building with AI and the results got good, I made the wrong inference. I thought AI had gotten better at design. What had actually happened was that I had gotten better at directing it — and I mistook the output for the intelligence.</p><h3>What Actually Happened When It Worked</h3><p>When I began building in Google AI Studio, I wasn’t handing AI a design problem. I was specifying every constraint, one by one. “Show this field only when trainer mode is selected.” “This role can see all entries; this role can only see their own.” “When the user completes this step, advance to this state, not that one.” Each prompt was a design decision I had already made — I was just using AI to execute it.</p><p>What surprised me was how faithfully it executed. When the direction was precise, it didn’t improvise. It didn’t average across what it had seen before. It just did exactly what I described. That’s when I realized: the quality of the output was a direct reflection of the quality of the direction. AI didn’t contribute design intelligence. It contributed reliable execution.</p><h3>Direction Is the Design</h3><p>This is the thing I had wrong. I thought design was the construction — the act of building the thing. So when AI started building it, I assumed design had transferred to AI. But construction was never the design. The design was every decision that preceded it: what the user needs, what states the system has, what logic governs the transitions, what the edge cases are, and why any of it matters.</p><p>None of that came from the AI. All of it came from me. The prompts weren’t shortcuts — they were the work. The effort it took to specify conditional logic precisely, to anticipate role conflicts, to describe what should happen in states I hadn’t built yet — that effort is design. AI just made it possible to test those decisions immediately, against something that ran.</p><p>The test is simple: if AI could design, the vague prompt would work. “Design a great onboarding flow” would be enough. It never is. The more effort it takes to direct AI toward something useful, the more that effort is proof of where the design actually lives.</p><h3>How the Workflow Changed</h3><p>What did shift — genuinely — was the sequence. I now start by building the functional skeleton before ever opening Figma. I work through logic, states, and conditional behavior first. Once that’s stable, I move into Figma to design the component library and system around it. The visual design reflects actual functionality instead of anticipating it.</p><p>That’s a real change. In the old workflow, you’d design screens that implied behavior, then hand them to a developer to interpret. Now the behavior is proven first, and the visual design gives it form. It’s a better order of operations — not because AI designed anything, but because AI made it cheap to test functional logic before committing to visual design.</p><h3>An Unexpected Layer: Designing the Conversation</h3><p>One thing I didn’t anticipate was having to design the AI chat experience inside the app itself. It became more than a utility — it acted as an interface to the data. Users could query, manage, and navigate functionality through language. That made the chat a living UI — one that required its own design thinking: what questions should it answer, what should it decline, how does it surface information without overwhelming the user?</p><p>Designing that interaction felt like shaping a dialogue between logic and user intent. Which is, when I think about it, just design — applied to a new surface.</p><h3>What AI Still Can’t Provide</h3><p>AI doesn’t conduct research. It doesn’t know what your users actually struggle with, what competitors are doing, or what business constraints are non-negotiable. It can’t decide which features matter or why the product should exist. It doesn’t care if the output serves anyone — it just produces what the prompt describes.</p><p>It also doesn’t notice its own mistakes. It follows conflicting instructions without flagging the conflict. It rebuilds something you already decided to remove without questioning why. The clarity about what matters, the restraint to stop when the direction isn’t right — that’s still entirely on the designer.</p><p>AI got better at executing design. That’s genuinely useful. But execution was never the hard part.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=1f97f60677b0" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Why AI can’t design (on its own)]]></title>
            <link>https://medium.com/@danniesnchez/why-ai-cant-design-on-its-own-603e20c2b5ae?source=rss-8f14711e9d01------2</link>
            <guid isPermaLink="false">https://medium.com/p/603e20c2b5ae</guid>
            <category><![CDATA[ui-design]]></category>
            <category><![CDATA[ux]]></category>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[design]]></category>
            <category><![CDATA[future-of-work]]></category>
            <dc:creator><![CDATA[Dannie Sánchez]]></dc:creator>
            <pubDate>Tue, 21 Oct 2025 23:42:33 GMT</pubDate>
            <atom:updated>2025-10-21T23:42:33.674Z</atom:updated>
            <content:encoded><![CDATA[<p>AI can come up with stunning videos, surreal animations, and even turn your raw data into a thorough, readable report in minutes.</p><p>So why can’t it design a digital product?</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*0M26ZGEQglOT_QJO" /><figcaption>Photo by <a href="https://unsplash.com/@marvelous?utm_source=medium&amp;utm_medium=referral">Marvin Meyer</a> on <a href="https://unsplash.com/?utm_source=medium&amp;utm_medium=referral">Unsplash</a></figcaption></figure><h3>It Feels Like It Should Be Easier</h3><p>Think about it: generating a superhero flying through a dense city takes advanced simulations, 3D physics, lighting models, and millions in special effects budgets. In comparison, an app interface?</p><p>It’s just rectangles and text, right?</p><p>So it makes sense to expect AI to handle product design.</p><p>In theory, it should be the easy part.</p><h3>But Design Isn’t Visual Output — It’s New Logic</h3><p>Here’s the catch: you’re not asking AI to remake something it’s seen.</p><p>You’re asking it to design a product that doesn’t exist yet — with logic it’s never encountered.</p><p>AI thrives when patterns repeat. But product design rarely repeats.</p><p>Every new product is a custom combination of flows, users, constraints, and intentions. And that specific mix? It’s probably not in the training data.</p><h3>How a Language Model Thinks</h3><p>Large language models don’t understand meaning.</p><p>They understand <em>what usually comes next</em>.</p><p>During training, these models are fed enormous volumes of text — books, websites, emails, help docs, even code. That text gets broken down into small units called <strong>tokens</strong> — think words, parts of words, or characters.</p><p>Each token is mapped into a space called <strong>embedding space</strong> — a kind of abstract coordinate system. You can imagine it like a massive map, or a star chart. Words that often appear near each other — like “cart,” “checkout,” and “shipping” — end up clustered in the same region.</p><p>The model doesn’t know what those words mean. It just knows they tend to orbit close together.</p><p>So when you give it a prompt, it doesn’t reason through what should happen next. It just <strong>predicts</strong> what would most likely come next, based on where you dropped the first word in this invisible galaxy of tokens.</p><p>That’s how it writes decent emails, bug reports, or code comments.</p><p>It’s playing the odds — not designing a system.</p><h3>Why Relational Meaning Doesn’t Work in UI Design</h3><p>In language, this works beautifully.</p><p>Words gain meaning from their neighbors. “Cart” lives near “checkout.” “Submit” appears often with “form.” The model learns patterns — not definitions, just proximity.</p><p>In images and video, it also works — a certain color, shape, or lighting pattern evokes context, mood, or style. AI can remix those effectively because the rules are flexible, and the output is about vibe.</p><p>But in product design, that logic collapses.</p><p>A button doesn’t mean something because of what it’s next to.</p><p>It means something because of what it <em>does.</em></p><p>You can’t infer design logic from visual adjacency.</p><p>Only the original designer knows why that modal opens after step 3, or why that field disappears when you select “enterprise.”</p><p>AI doesn’t know the flow. It just sees proximity.</p><p>And proximity, in this case, tells it nothing useful.</p><h3>You Can’t Remix Your Way Into a Real Product</h3><p>AI is great at visual invention because it’s remixing what it already knows.</p><p>Give it the word “sign,” and it pulls from thousands of examples of street signs. Their shapes, colors, materials, lighting angles. It doesn’t recreate any one of them. It blends them. Synthesizes. That’s what diffusion models do — they recreate vibes from noise.</p><p>It works beautifully for visuals. You get something that <em>looks</em> like a sign.</p><p>But products aren’t just something you look at. They’re something you move through.</p><p>You can’t remix flow logic. You can’t collage together role hierarchies. There’s no visual shorthand for why this button leads to that screen, or why that field disappears when a dropdown is selected.</p><p>Product design isn’t about the surface. It’s about what the surface does.</p><p>And what it does depends on what came before, what’s happening now, and what constraints you’re solving for. No training data set has <em>your exact</em> mix of users, roles, logic, and business need.</p><p>Remix doesn’t help. It gets in the way.</p><h3>Prompting Isn’t Precision</h3><p>When people talk about prompting AI to design, they assume it works like giving instructions to a human. But it doesn’t. A prompt isn’t interpreted — it’s encoded.</p><p>It gets converted into a <strong>dominant embedding</strong> — a mathematical fingerprint of your words — surrounded by a cloud of related sub-embeddings. These are positioned in vector space, not for their literal meaning, but for how they statistically relate to other data the model has seen.</p><p>The result? Even one word can shift the entire output.</p><h3>The Illusion of Similarity</h3><p>Even when an AI system has been trained on a wide range of apps in a given category — like habit trackers — <strong>each example represents slightly different design decisions</strong>: in layout, hierarchy, functionality, and tone.</p><p>These differences aren’t random. They’re purposeful responses to specific business models, user behaviors, technical constraints, and visual identity systems. But the AI doesn’t see those reasons — it only sees patterns.</p><p>When prompted to “design a habit tracker,” the model statistically averages the most common features and structures across those examples. The result is a <strong>softmax of the category</strong> — the center of mass, not the edges.</p><p>As the prompt moves further from common examples — toward niche use cases or novel formats — the training data thins. And the outputs don’t become more tailored. They become more generic.</p><p><strong>The divergence increases</strong>, not in a linear or interpretable way, but in ways that can feel <strong>arbitrary</strong> or <strong>incoherent</strong>, especially when trying to solve for edge-case needs.</p><h3>Every Word Tilts the Result</h3><p>This is the hard part to grasp: prompts are not additive. They’re spatial.</p><p>Every adjective, every noun, even the order of your words affects the location in embedding space. If something contradicts, overlaps, or introduces ambiguity — the model doesn’t reconcile it. It just lands somewhere between the fragments and improvises.</p><p>That might be fine in video or image generation. A little variance adds flavor.</p><p>But in product design? Those tiny shifts aren’t aesthetic. They’re <strong>structural</strong>.</p><p>One misplaced word can change role logic. One missing constraint can drop a flow.</p><p>A single contradiction can ripple through the whole interface.</p><h3>Design Isn’t Description — It’s Decision</h3><p>Design isn’t a list of features.</p><p>It’s a system of interlocking decisions: what appears, when, for whom, under what constraint, with what consequence.</p><p>You can’t prompt that.</p><p>You can describe what your app should do — “users can track workouts and share results with friends” — but you’re not describing <em>how</em> that logic unfolds across states, roles, edge cases, or interactions.</p><p>And that’s what design is.</p><p>A prompt might help sketch a UI.</p><p>But it can’t hold the structure in place. It can’t see the tension between screen 2 and screen 12. It can’t enforce logic across flows.</p><p>Because design isn’t what you say.</p><p>It’s what you decide.</p><h3>Design Is Intentional — And AI Doesn’t Know Why</h3><p>Design isn’t decoration — it’s an <em>intentional solution</em> to a problem or task. It involves trade-offs, context, and user behavior, not just visuals.</p><p>Even if AI sees patterns across thousands of habit trackers, each one makes slightly different decisions in layout, hierarchy, and flow — reflecting real human goals and constraints. As apps diverge in purpose or audience, those small differences compound into major ones.</p><p>The further you drift from the core examples in its training data, the more AI flattens outcomes into generic averages.</p><p>The results look plausible — but they miss the why.</p><p>And in design, <em>why</em> is everything.</p><h3>Figma Is Close — but Not Quite There Yet</h3><p>Figma has introduced AI features that speed up early design, like autogenerating layouts or naming layers. It also supports design systems through emerging tools like the Model Context Protocol (MCP), which lets AI reference component libraries and tokens. But it doesn’t yet enforce strict design constraints, flag deviations, or audit ripple effects across a system when new components are introduced.</p><p>What’s missing is the rigor: a feedback loop where AI proposes components only using approved primitives, checks for contradictions, and learns incrementally through human edits. That vision is forming — but it’s not fully realized yet.</p><h3>The Real Limits of AI in Design</h3><p>AI design tools are great for early-stage momentum — they help scaffold a new product or bootstrap a basic design system. They move fast when flexibility is high, ambiguity is tolerated, and the stakes are low. But as the product evolves, so do its constraints.</p><p>New edge cases arise. Existing rules get bent or broken. The design system expands in ways that aren’t cleanly hierarchical — they ripple across templates, patterns, and decisions made months earlier.</p><p>Without understanding <em>why</em> something was designed a certain way, AI can’t evolve the system. At best, it starts to contradict itself. At worst, it reinforces outdated patterns. And without clear source-of-truth rules and human judgment, it lacks the ability to audit and adapt responsibly.</p><h3>So as the product gets more mature, AI gets lost in the weeds</h3><p>AI thrives in ambiguity, but mature products demand precision. Early designs can tolerate fuzzy logic and pattern-matching. But once users rely on it, every detail matters. Decisions cascade. Edge cases multiply. What once felt like creativity becomes constraint management.</p><p>Without memory of why each choice was made, AI can’t adapt responsibly — it reverts to remixing. And remixing is no longer enough. Because what matters now isn’t what the product looks like. It’s how the whole thing holds together.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=603e20c2b5ae" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[The House That AI Drift Built]]></title>
            <link>https://medium.com/@danniesnchez/the-house-the-ai-drift-built-b94b11099237?source=rss-8f14711e9d01------2</link>
            <guid isPermaLink="false">https://medium.com/p/b94b11099237</guid>
            <category><![CDATA[product-design]]></category>
            <category><![CDATA[artificial-intelligence]]></category>
            <category><![CDATA[design-tools]]></category>
            <category><![CDATA[ai-and-design]]></category>
            <category><![CDATA[user-experience]]></category>
            <dc:creator><![CDATA[Dannie Sánchez]]></dc:creator>
            <pubDate>Tue, 29 Jul 2025 17:17:58 GMT</pubDate>
            <atom:updated>2025-08-16T21:06:09.606Z</atom:updated>
            <content:encoded><![CDATA[<p>Everyone’s talking about how AI can now “design products.” Tools like Lovable promise full interfaces from a prompt: “Build a contact form with admin access” or “Make a simple onboarding flow.” It’s fast. It feels like a shortcut.</p><p>But here’s the catch: AI is great at recreating patterns it’s seen before. The moment you ask for anything <strong>slightly contextual</strong> — like role-specific access, conditional fields, or form logic — it forgets. It’ll ignore your role requirements on one screen, then invent new ones you never asked for. What looks usable at first becomes a web of assumptions you didn’t authorize.</p><p>That’s the real problem: <strong>AI improvises. Design can’t afford improvisation.</strong></p><p>In art, it adds flavor. In product design, it adds chaos.</p><p>This piece breaks down why tools like diffusion models and LLMs fall apart when asked to lead design — and how to flip the relationship: <strong>human-led, AI-assisted.</strong></p><h3>Diffusion Models: Stylists Without Structure</h3><p>Diffusion models — like the ones powering Midjourney or DALL·E — are impressive at visual style. Ask for “a brutalist hero section with neon typography” and you’ll get something that <em>looks</em> like design. But it’s not design — it’s decoration. These models don’t understand layout logic — they imitate layout patterns. They’ve seen enough websites to recognize a hero section, but they don’t know <em>why</em> certain elements go where they do. There’s no awareness of user intent, and no memory of what came before.</p><p>They don’t know where a button should go, or why white space matters — they just generate what <em>looks</em> plausible. The results feel consistent because they mimic familiar compositions. But as soon as you need something tied to your brand, logic, or user flows, they drift.</p><p>That’s because diffusion models don’t design. <strong>They collage.</strong> And in product design, copy-paste aesthetics aren’t enough. Real UI needs cohesion, hierarchy, interaction, and purpose.</p><h3>LLMs: Logical, but Not Visual</h3><p>Large Language Models — like the ones behind Lovable or Framer AI — excel where diffusion models fail: logic. They’re much better at generating consistent naming, CRUD flows, user journeys, and even component structures. Ask it to build a screen for logging workouts, and it’ll return the right fields, labels, and maybe even the correct order.</p><p>But it can’t <em>see</em>. It doesn’t know if the layout feels balanced, if the spacing is off, or if the visual hierarchy is broken. It might give your primary button too little weight, misalign your heading sizes, or place elements in a way that breaks the flow. It doesn’t know it’s done that — because it doesn’t reason visually. It just predicts what comes next in text.</p><p>Worse, LLMs have short memories. They can forget prior decisions, like field requirements or user roles. They’ll invent labels, rename flows midstream, or offer slight variations on the same component — introducing design drift in slow motion.</p><p>They’re smart, but not self-aware. And in design, awareness is everything.</p><h3>Drift: AI’s Silent System Killer</h3><p>Drift isn’t loud. It doesn’t crash your layout or throw an error. It shows up quietly — screen by screen, prompt by prompt — as small inconsistencies that compound until your product loses all cohesion.</p><p>One screen calls it “Create Log,” the next says “New Entry.” A button is rounded on one page and sharp on another. Your user roles start as “Admin” and “User,” but somehow become “Manager,” “Owner,” and “Superuser” as the AI improvises. None of these choices are wrong individually — but together, they create a product that feels disjointed and unreliable.</p><p>That’s the danger: <strong>AI doesn’t remember decisions unless you force it to</strong>. It generates in isolated guesses, not connected systems. So when you rely on it to build screen after screen, it copies the last version imperfectly. Then it copies the copy. Then it copies that.</p><p>Drift is fine in a one-screen demo. In a full product, it’s a structural collapse.</p><h3>Drift Doesn’t Scale — It Compounds</h3><p>Drift in a one-off design isn’t a problem. In art, it’s part of the charm. If a graffiti piece has asymmetry or the colors shift slightly — it’s fine. No one expects structural logic in a standalone creative work.</p><p>But design isn’t one-off. It scales. And as it scales, drift doesn’t just <em>accumulate</em> — it <strong>compounds</strong>.</p><p>It’s like building a one-story house slightly off-center. The floor’s a little uneven, the doorframes don’t quite match — but it stands. Now build a 30-story apartment complex on top of that crooked base. Every misalignment gets magnified. The plumbing doesn’t connect. The walls don’t meet. The elevator shaft doesn’t line up across floors. What was tolerable in a tiny space becomes catastrophic at scale.</p><p>That’s exactly what happens when you let AI improvise screen after screen. It doesn’t just make small mistakes — it repeats, mutates, and amplifies them. <strong>Drift isn’t a cosmetic issue. It’s exponential system failure.</strong></p><h3>Art Welcomes Improvisation. Design Can’t.</h3><p>In art, improvisation is freedom. It’s what makes graffiti wild, abstract painting visceral, or jazz unpredictable. No one questions a shifted line or uneven rhythm. It’s expression.</p><p>But design is not expression. It’s structure. It’s function. It’s timing. And improvisation in design doesn’t come across as creative — it comes across as broken.</p><p>When an AI drifts from the type scale, invents a new shade of blue, or mislabels a button, it’s not exploring — it’s deviating. It’s one violinist in the orchestra deciding to play a different tempo. It doesn’t add life — it throws the whole system out of tune.</p><p>Design is orchestrated. It works because everything plays in sync. Once AI is allowed to freelance, even slightly, the user experience begins to fray. Not visually — structurally.</p><p>That’s the core misunderstanding: AI thrives when variation is allowed. But design punishes variation. And that’s why giving AI creative freedom in design isn’t innovation — it’s entropy.</p><h3>The Role of AI: Assistant, Not Architect</h3><p>AI isn’t useless. In fact, it’s remarkably good at certain tasks — just not the ones people keep asking it to do.</p><p>It’s not a design lead. It’s not a strategist. It’s not an architect. But it can be a sharp, tireless assistant.</p><p>Ask it to generate CRUD scaffolding, placeholder copy, or variations on a layout you’ve already defined? It flies. Ask it to review your Figma file for inconsistent padding, rogue color tokens, or broken component instances? It’ll catch things even your lead designer might miss.</p><p>That’s the irony: the same AI that drifts when it builds is <em>brilliant</em> at spotting drift when it reviews. It improvises poorly, but audits with precision.</p><p>Used this way — under structure, with intent — AI becomes powerful. It doesn’t replace the designer. It extends the designer’s reach.</p><p>The future isn’t AI-led design. It’s designer-led systems, supercharged by AI.</p><h3>Conclusion: Design Needs Intent, Not Improv</h3><p>AI can generate faster than any human. But it doesn’t know why it’s generating. It can mimic patterns, guess layouts, and suggest flows — but it has no stake in whether those decisions make sense, scale, or serve real people.</p><p>That’s the difference. Design isn’t about speed. It’s about <strong>intent</strong>. About building systems where every choice connects to a purpose. Where consistency builds trust. Where flow matters as much as form.</p><p>When AI improvises in art, it creates magic. But when it improvises in design, it creates noise. The tools aren’t the problem. The framing is. If you treat AI as the lead, you’ll spend your time untangling decisions it made blindly. But if you lead — and let AI assist — you get the best of both: speed without sacrificing clarity.</p><p>The future of design isn’t autonomous. It’s <strong>orchestrated</strong> — by humans, with AI as part of the ensemble.</p><h3>What Do Now?</h3><p>Ironically, I didn’t learn AI’s design limitations by using it to design. I learned them by using it elsewhere — writing code, analyzing ideas, streamlining my photography workflow. It helped in many places. But when structure mattered, it drifted. It improvised. And that’s where it broke down.</p><p>For me, AI has been most useful in design when I use it to talk through logic, structure, and copy — not to create finished products. When I lead, it helps. When I let it lead, I usually regret it.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=b94b11099237" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[7 Practical Ways to Use ChatGPT to Boost Your Product Design Workflow: Part 2]]></title>
            <link>https://medium.com/@danniesnchez/7-practical-ways-to-use-chatgpt-to-boost-your-product-design-workflow-part-2-25d1101fd10e?source=rss-8f14711e9d01------2</link>
            <guid isPermaLink="false">https://medium.com/p/25d1101fd10e</guid>
            <category><![CDATA[ai-tools]]></category>
            <category><![CDATA[design-research]]></category>
            <category><![CDATA[ux-design]]></category>
            <category><![CDATA[ai-in-design]]></category>
            <category><![CDATA[product-design]]></category>
            <dc:creator><![CDATA[Dannie Sánchez]]></dc:creator>
            <pubDate>Tue, 03 Jun 2025 01:00:39 GMT</pubDate>
            <atom:updated>2025-06-03T01:00:39.070Z</atom:updated>
            <content:encoded><![CDATA[<p>In the previous article, we explored how AI can support product design through learning, idea generation, and user research. In this second part, we’ll cover three more ways AI can refine your process, focusing on design systems, prototyping, and user documentation. You’ll see how to stay in control while using AI as a partner in your work.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*-kYTcZmiJuvBDzRdqXFfdQ.png" /><figcaption>Careful direction throughout the product design process</figcaption></figure><h3>4. Design System Refinement</h3><p>Design system creation is a time-consuming process. You have to carefully make design decisions, working from brand colors (or inherited brand guides), standard type sizes, and the various uses you need to cover. Typically, this involves tedious, repetitive tasks like populating tokens and organizing styles. Figma has decided to ignore this obvious part of product design, so we’re left to figure it out ourselves.</p><p>GPT can assist with the repetitive parts, but remember that we are the ones making the decisions. AI doesn’t have ‘taste’ or an understanding of context and human nuances, so you still need to guide the process.</p><p><strong>Example prompt to try:</strong></p><ul><li>Refine this brand color palette for a cohesive and balanced visual system using a harmonious color scheme (like complementary, analogous, or split-complementary). Create system colors (error, success, warning, info, background, text) that are distinct yet cohesive. Provide hex codes and clear, human-readable names for each. <strong>[paste brand colors]</strong></li></ul><p>Sometimes a prompt will work, but other times, a <strong>meta-prompt</strong> is needed. A meta prompt is a higher-level instruction that asks an AI chatbot to generate a detailed prompt for you, rather than an immediate answer. It’s a way to co-create the structure of your prompts so you get more tailored and powerful results.</p><p><strong>Example meta-prompt to try:</strong></p><ul><li>Using a modular ratio of <strong>[1.25]</strong>, create a scalable type system for a digital product using <strong>[font name]</strong> with 4 heading sizes and 2 paragraph sizes. Use a consistent naming convention (like fontSizeHeading1). Calculate REM values based on a base font size of 16px, and suggest line heights for accessibility. Present the result as a table with token name, pixel size, REM value, and recommended line height for clarity.</li></ul><h3>5. Designing for Collaboration</h3><p>Managing a design project can become complicated as assets accumulate, making it challenging to keep everything in order. Providing clear guidance to developers is crucial during implementation. ChatGPT can help by offering structured advice on categorizing and documenting your assets. While it can’t directly interact with your design tools, you can describe your needs and get recommendations on how to organize your files. By leveraging ChatGPT’s capabilities, you can offload some of the tedious work and maintain a well-structured design file.</p><p><strong>Example meta-prompt to try:</strong></p><ul><li>Explain how atomic design relates to how design assets are organized, specifically in the context of forms and form fields.</li><li>Explain how using the BEM naming convention can help organize Figma design assets and create a smoother designer-to-developer handoff.</li></ul><p>Ultimately, this ensures your design system remains a collaborative, well-documented asset that supports efficient workflows.</p><h3>6. Product Strategy</h3><p>Developing a robust product strategy is crucial for any digital product, but organizing the insights from team workshops and user interviews can be overwhelming. ChatGPT can help by acting as a second set of eyes. It sorts through your notes, sticky screenshots, and raw data to create logical themes and groupings. In my own work, I had a massive collection of stickies in Figma, and by sharing screenshots with GPT, it helped me organize them into more structured ideas and themes for easier tracking. While you’re still the one making the final decisions, GPT can lighten the load of managing this early-stage data.</p><p><strong>Example prompts to try:</strong></p><ul><li>I have a collection of sticky note insights from a user research workshop. Each note contains a short statement from a participant. Please do two things:<br>1. Organize them into rows or groups by similarity to identify how many times a topic was mentioned.<br>2. Then, cluster these groups into broader themes (e.g., pain points, communication issues, technical issues).</li><li>Take a look at this customer profile in the Value Proposition Canvas. Based on the customer jobs, pains, and gains provided, suggest ideas for the left side of the canvas: features, services, and solutions that could create strong product-market fit. Ensure the suggestions are relevant to the design context and directly address the noted pain points and goals of the user group.</li><li>Take the following product feature suggestions [paste/add file] and organize them into an impact/effort matrix. Each feature should be placed in one of four quadrants: Quick Wins (high impact, low effort), Major Projects (high impact, high effort), For a Rainy Day (low impact, low effort), and Time Wasters (low impact, high effort). Provide a clear explanation for why each feature belongs in that quadrant.</li></ul><p>To summarize, this second part explored three more ways to use AI as a partner in your product design workflow: refining your design system, organizing your design files for smoother collaboration, and crafting a product strategy that aligns with user insights. While AI can’t replace human intuition, it can accelerate these foundational tasks so you can focus on the bigger picture. Stay tuned for the final part of the series, where we’ll wrap up with a few more practical ways to integrate AI into your work.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=25d1101fd10e" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[7 Practical Ways to Use ChatGPT to Boost Your Product Design Workflow: Part 1]]></title>
            <link>https://medium.com/@danniesnchez/7-practical-ways-to-use-chatgpt-to-boost-your-product-design-workflow-a8f83673831d?source=rss-8f14711e9d01------2</link>
            <guid isPermaLink="false">https://medium.com/p/a8f83673831d</guid>
            <category><![CDATA[product-design]]></category>
            <category><![CDATA[ux-design]]></category>
            <category><![CDATA[design-research]]></category>
            <category><![CDATA[ai-in-design]]></category>
            <category><![CDATA[ai-tools]]></category>
            <dc:creator><![CDATA[Dannie Sánchez]]></dc:creator>
            <pubDate>Mon, 02 Jun 2025 01:29:06 GMT</pubDate>
            <atom:updated>2025-06-02T18:26:22.197Z</atom:updated>
            <content:encoded><![CDATA[<p><em>This article explores how you can use ChatGPT to learn faster, generate ideas, and streamline your product design workflow. Here are seven real-world applications for designers.</em></p><blockquote>Before diving into these seven applications, here’s a quick review of what ChatGPT is and what it is not. ChatGPT is an AI-powered tool that generates text responses based on patterns in its training data. It can provide helpful suggestions, organize information, and explain concepts, but it does not create new knowledge or understand context in the same way a human would. Use it as a starting point and an assistant that is always ready to help refine your Product Design work.</blockquote><p>AI will not write the next great novel, create the breakthrough marketing strategy, or come up with the next big product idea. You will and can, with the assistance of AI. To fully tap into this potential, let go of the idea that AI is here to take the wheel while you watch. Instead, see it as a collaborative partner that can generate ideas, spark insights, and streamline your workflow, all while still needing your direction and judgment to make them shine.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*9UH8_DovaG5R3QRCuDNzuQ.png" /><figcaption><em>That’s you, on the left. Don’t let little GPT take the wheel.</em></figcaption></figure><h3>1. Learning and skill development</h3><p>Imagine you’re a product designer unfamiliar with design frameworks like Design Thinking or the Value Proposition Canvas. You know they exist, but you’re not sure how they work or how they might complement each other in your process. AI chatbots can act as your personal tutor, explaining these frameworks, comparing them, and helping you see how they can support your product design work.</p><p><strong>Here are some example prompts you could try with your AI chatbot to explore this concept further:</strong></p><ul><li>Provide a brief overview of what <strong>[Framework]</strong> is and how it applies to product design.</li><li>Explain <strong>[term]</strong> within <strong>[framework]</strong> and how it contributes to <strong>[stage in product design]</strong>.</li><li>Compare the strengths and limitations of <strong>Design Thinking</strong> and the <strong>Value Proposition Canvas</strong>.</li><li>Suggest ways to combine <strong>Design Thinking</strong> and the <strong>Value Proposition Canvas</strong> to get the most from each.</li><li>Explain how each stage of <strong>[methodology]</strong> contributes to the final product design outcome.</li></ul><h3>2. Inspiration &amp; idea generation</h3><p>Following up on ChatGPT’s strongest feature, I want to address one of its perceived advantages that can actually be a disadvantage. It might seem like a major benefit that anyone can generate a great idea with a simple prompt, but that’s misleading. AI can’t create something truly new or original. It’s just recycling patterns from what it was trained on. While it can help you brainstorm quickly, it can’t replace the unique insights, experience, and judgment that skilled product designers bring. Ultimately, the advantage belongs to those who know how to turn raw ideas into real breakthroughs.</p><p>What we actually get is a variety of ideas — some cliché, some practical, some impractical, and occasionally some that might be worth exploring further. ChatGPT has an endless supply of these, but it’s up to <strong>you</strong> to filter out the noise and find what’s truly useful.</p><p>Unlike learning new concepts, using AI for idea generation is not about expecting the AI to take the wheel and hand you the perfect idea. Instead, it is most effective when <strong>you</strong>, the product designer, are in control and ask the AI to analyze and learn from real-world examples that are already working. By studying what is successful, you can find inspiration and insights that help you shape something new and meaningful.</p><p><strong>Example prompts to try:</strong></p><ul><li>Provide a brief analysis of why <strong>[specific product or feature]</strong> is considered a good example of product design.</li><li>What features or approaches in <strong>[product name]</strong> might be adapted to improve a <strong>[similar/different product]</strong>?</li><li>Provide a short explanation of a known feature or design decision in <strong>[product or feature]</strong> and a possible drawback or limitation associated with it.</li><li>Identify the core user needs addressed by <strong>[product or feature]</strong> and how these could inspire new features for <strong>[specific audience, context, or niche]</strong>.</li></ul><h3>3. User Research</h3><p>User research is a key part of product design, but it often involves combing through large volumes of both structured and unstructured data. This is where AI chatbots like ChatGPT shine. They can quickly process and summarize insights, helping you spot themes and patterns faster. Once you know what insight you’re trying to gather, you can go through each research question one by one or get all the key insights listed out at once. AI excels at these repetitive tasks, synthesizing and refining the data to the desired level of granularity.</p><p>However, you cannot go on auto-pilot. AI can help, but it cannot truly lead the way. It is up to you to stay in control, ensuring that the insights and direction still come from human judgment. While AI won’t replace the empathy and nuance of human-led research, it can dramatically speed up the analysis process and help you focus on interpreting the findings and making better product decisions.</p><p><strong>Example prompts to try:</strong></p><ul><li>Rephrase these team-provided user research questions to be unbiased and suggest the best format (open-ended, scale, multiple choice) for each: <strong>[paste questions]</strong>.</li><li>Group these user research questions into logical categories to create a clear interview structure: <strong>[paste questions]</strong>.</li><li>Create a step-by-step proto-user journey for <strong>[user task or goal]</strong> that can identify potential pain points and inform user interview question development.</li><li>Count how many times <strong>[specific topic or feature]</strong> was mentioned in this user research data: <strong>[paste data]</strong>.</li><li>List any repeated phrases or topics in this user research data without interpretation: <strong>[paste data]</strong>.</li></ul><p>To briefly summarize, in this first part we explored three practical ways to use AI chatbots: learning and skill development, idea generation through analyzing existing examples (not relying on AI for magic-bullet ideas), and organizing and summarizing user research data. In the next article, we’ll cover three more ways to incorporate AI into your product design work.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=a8f83673831d" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>