<?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 Clare Cady on Medium]]></title>
        <description><![CDATA[Stories by Clare Cady on Medium]]></description>
        <link>https://medium.com/@clarelcady?source=rss-6d4e528f05b2------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*ZcJBI08E0_FU-iJtDt4HvA.png</url>
            <title>Stories by Clare Cady on Medium</title>
            <link>https://medium.com/@clarelcady?source=rss-6d4e528f05b2------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Sat, 21 Nov 2026 03:34:20 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@clarelcady/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[Vibe Coding Isn’t Dead]]></title>
            <link>https://medium.com/@clarelcady/vibe-coding-isnt-dead-062f87de1f6a?source=rss-6d4e528f05b2------2</link>
            <guid isPermaLink="false">https://medium.com/p/062f87de1f6a</guid>
            <category><![CDATA[technology]]></category>
            <category><![CDATA[founders]]></category>
            <category><![CDATA[diversity-and-inclusion]]></category>
            <category><![CDATA[vibe-coding]]></category>
            <category><![CDATA[startupş]]></category>
            <dc:creator><![CDATA[Clare Cady]]></dc:creator>
            <pubDate>Tue, 27 Jan 2026 13:06:16 GMT</pubDate>
            <atom:updated>2026-01-27T13:07:03.968Z</atom:updated>
            <content:encoded><![CDATA[<p>You just don’t see who it benefits</p><figure><img alt="Image of an open with colforul keys sitting on a desk. There is a black screen with code written on it. Beside it is a water bottle covered with colorful stickers." src="https://cdn-images-1.medium.com/max/1024/1*DL19EyKzG2YyY1RYkUYoww.jpeg" /></figure><p>When vibe coding first hit the mainstream, it came with all the hype tech loves most: that it is the new disruption that changes everything.</p><p>Anyone can build software! Engineering is dead now! Update your resumes!</p><p>People were laid off across the PDLC. Who needs this many engineers, product managers, scrum masters, UX designers, or UX researchers, if you can produce these products with some well-crafted sentences? I myself fell victim to this phenomenon — laid off along with 60% of my team. We saw record unemloyment in tech. Sure it’s not <em>just </em>the fault of vibe coding. Other AI tools, a cooling market, and political factors also contributed. But vibe coding tools got deep into the core of what it means to build software. When engineering started to fall, it took out a lot of other dominoes.</p><p>But, by the end of 2025, the tone shifted. Suddenly vibe coding was reckless. It was immature. It was a shortcut that produced fragile systems no one could maintain. The same people who had framed it as a revolution began declaring it dead on arrival.</p><p>But here the thing, <em>both</em> narratives missed the point. Vibe coding didn’t fail. It was simply aimed at the wrong problem.</p><h3>The Original Promise (and the Original Mistake)</h3><p>The early mistake was not technical, it was conceptual. Vibe coding, the use of LLMs to generate applications with verbal prompts, was framed as a replacement for engineering rather than a complement to it. The issue right out the gate is framing in that way guaranteed disappointment. Software engineering isn’t just about producing code — it’s about maintaining systems over time, managing risk, handling scale, and navigating edge cases that only show up once real humans start using real products.</p><p>AI tools are not good at that kind of stewardship. Not yet. It requires a human in the loop to sustain the work over time.</p><p>As startups rushed to ship AI-generated tools, the same problems surfaced again and again. Codebases worked until they didn’t. Small feature changes caused cascading failures. Debugging became nearly impossible because no human had authored the system in the first place… meaning no uhman really understood it. Tech debt — and UX debt — accumulated faster than teams could pay it down. Other issues included:</p><ul><li>AI-generated code often lacks consistent architecture, meaningful documentation, and long-term maintainability unless heavily reviewed by experienced engineers.</li><li>Models could generate code that <em>looked</em> correct while quietly introducing vulnerabilities.</li><li>The products produced often lacked the understanding of human behavior and needs. Products were not adoption-ready because they were not designed with human problems in mind.</li></ul><p>None of this means vibe coding is inherently dangerous. It means it behaves like every other powerful tool: it magnifies both good judgment and bad judgment.</p><h3>So, it’s dead now?</h3><p>Once these limitations became visible, the pendulum swung hard in the opposite direction. Articles and threads started declaring vibe coding overhyped or effectively useless. Some went further, arguing that the entire category had been a distraction from “real” engineering work. Usage and valuation of tools like Lovable and Replit plummeted.</p><p>But here’s the thing. That backlash relied on the same flawed assumption as the hype: that vibe coding’s value should be measured by whether it can replace engineers.</p><p>I think that’s the wrong benchmark.</p><h3>What Vibe Coding Is Actually Good At</h3><p>Quietly, away from the loudest discourse, teams began using vibe coding in less extreme ways. Not as a substitute for engineering, but as an accelerator for early-stage work.</p><p>Vibe coding turns out to be extremely effective for prototyping, exploration, and validation. It allows founders and teams to move from idea to tangible artifact quickly, test assumptions with users, and discard bad ideas before they become expensive. It also makes communication with technical teams easier — prototypes become living specifications instead of abstract documents.</p><p>In this role, vibe coding doesn’t need to be perfect. It needs to be <em>fast enough</em> and <em>good enough</em> to support learning and the option to launch proof of concept effectively.</p><p>And here’s where the data becomes interesting.</p><p>Recent reporting suggests that roughly a quarter of startups in some accelerator cohorts now use AI-generated code for the vast majority of their early codebases. At the same time, surveys consistently show that while a large majority of developers use AI tools in their workflow, far fewer are comfortable deploying AI-generated code without human review.</p><p>That gap matters. It shows that vibe coding hasn’t replaced expertise — it has reshaped <em>who gets to start</em> and <em>how quickly they can learn</em>.</p><p>And yes, there <em>are</em> successful launches out there using only vibe coded products. This isn’t an all or nothing conversation.</p><h3>So, who gets to start?</h3><p>This is the part that gets lost when vibe coding is evaluated solely through an engineering lens.</p><p>For non-technical founders, vibe coding collapses the distance between having an idea and having something real. It makes it possible to build a working prototype without first securing a technical cofounder, paying an agency, or spending months learning to code well enough to get started.</p><p>That matters because the traditional founder path assumes access — access to engineers, to capital, to networks that make early collaboration possible. Many people never get to the starting line not because their ideas are weak, but because the barrier to entry is too high.</p><p>This especially matters for underrepresented founders.</p><p>Women, founders of color, people with disabilities, LGBTQ folx, and others historically excluded from technical networks are disproportionately affected by the high startup friction. When you require investment and a technical cofounder before an idea can even be tested, you privilege people who already have proximity to engineering talent and VCs.</p><p>Vibe coding doesn’t solve structural inequality. But it does chip away at one very specific gate: the requirement that someone else must build your idea before it’s allowed to exist.</p><p>That’s not trivial. It changes who gets to try.</p><p>What’s exciting here is that not only does this grant greater entrance into the startup space for these types of founders, it opens up the possibility of addressing the problems they are uniquely positioned to see and solve. This has tremendous power to improve peoples’ lives in the areas of accessibility, equity, and visibility. It also opens up new markets that have not been reached — something many VCs are not considering. Apps focused on pelvic floor health, or helping people with ADHD become more organized, are coming from these founders. They have quietly and intentionally been applying the power of vibe coding to their issues, and delivering solutions that help people feel seen and supported.</p><h3>How did the hype miss this?</h3><p>Much of the discourse around vibe coding, the hype <em>and</em> the backlash, has been framed through a narrow lens: a senior, well-resourced, often white and male perspective that asks, “Does this replace me?”</p><p>From that vantage point, vibe coding is either terrifying or disappointing.</p><p>From this other vantage point, it’s neither. It’s enabling.</p><p>When tools are evaluated only by how they affect those already inside the system, their broader social impact is easy to miss. Vibe coding looks very different when viewed through an intersectional lens that accounts for unequal starting points and uneven access to opportunity.</p><h3>Vibe Coding Has a Future. It’s Just Not the One We Were Sold</h3><p>Vibe coding was never going to eliminate the need for skilled engineers. As with all AI tools, they are tools. They work best with a human in the loop that brings judgment, taste, and an understanding of the world that an LLM just does not have.</p><p>That idea was always more marketing than reality.</p><p>What it <em>can</em> do — and is already doing — is expand access to the act of building. It allows more people to test ideas, learn by doing, and enter the founder space without clearing every traditional hurdle upfront.</p><p>And that’s not dead. That’s just getting started.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=062f87de1f6a" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Confessions of an AI Founder Pt 1: what is “Founder” even?]]></title>
            <link>https://medium.com/@clarelcady/confessions-of-an-ai-founder-pt-1-what-is-founder-even-d85b2c998a93?source=rss-6d4e528f05b2------2</link>
            <guid isPermaLink="false">https://medium.com/p/d85b2c998a93</guid>
            <category><![CDATA[female-founders]]></category>
            <category><![CDATA[technology]]></category>
            <category><![CDATA[minority-founders]]></category>
            <category><![CDATA[confessions]]></category>
            <category><![CDATA[founders]]></category>
            <dc:creator><![CDATA[Clare Cady]]></dc:creator>
            <pubDate>Tue, 06 Jan 2026 17:31:14 GMT</pubDate>
            <atom:updated>2026-01-06T17:31:14.449Z</atom:updated>
            <content:encoded><![CDATA[<p>When do you cross the line into founderdom?</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*bko3n9TZwQQfdD68LJ_L1g.png" /></figure><p><em>My hands hold a gin and tonic with way too heavy a pour for my taste. The napkin wrapped around the glass is soaked with condensation. I stand in a conference room full of people chatting with one another, one of whom is giving me their elevator pitch. They are building an AI-powered product (I won’t say what it does), and are excitedly sharing with me their vision for its functionality and the people it will help. It’s amazing, and I am so grateful they had this idea and are willing to shepherd it into the world.</em></p><p>Then they ask me the question where I’m struggling.</p><p><em>“So, what are you working on?”</em></p><p>It’s an easy question to answer when I am wearing my consulting hat. I help founders like this person realize their vision with research and product strategy. But tonight I am not here to pitch my services. I am here as a peer. Someone who also has an idea that I am trying to realize.</p><p><em>“I’m a founder too… I guess.”</em></p><p>I don’t know why I worry about this answer. It’s totally true. I am building an AI-powered technology solution to a human problem. It was my idea (along with a co-founder), and I am doing the work 0–1. But when the words “I am a founder” sit on my tongue I worry — should I be calling myself that?</p><p>I don’t have a product to show.</p><p>I don’t have a pitch deck.</p><p>I don’t have any funding.</p><p>Or customers.</p><p>I have an idea, user data, and a Replit account with some APIs in a shell.</p><p>I know that this means I actually have quite a lot. I spend time mentoring people like me. Some of them don&#39;t have anything but the idea and vision. A lot of them don&#39;t have data. A lot of them don&#39;t have a pitch deck. A lot of them aren&#39;t sure what they&#39;re supposed to be doing, which is why they&#39;re talking to me.</p><p>This type of imposter syndrome is just so common in the founder space. It&#39;s extremely common for women, people of color, people with disabilities, and LGBTQ folx. The decision to call yourself a founder implies certain levels of knowledge, insight, and action. Or at least… It feels that way? I&#39;m not even exactly sure when I decided to call myself a founder, but I often wonder what that threshold is. Is a founder someone who intends to come up with an idea? Does the idea already need to be there? Is there a need for action? Or, is the concept of a founder someone who has built something?</p><p>In an industry where “fake it till you make it” is a huge part of “making it,” it seems that the sooner one can call themselves a founder, the better. The audacity to take on the identity early in the process may be part of the fire that gets lit that makes that process successful.</p><p>There&#39;s a lot of gatekeeping around founderdom. I think that&#39;s where the imposter syndrome can come from. If you look at LinkedIn profiles, you can see that someone may list themselves as a three-time founder… five-time founder… serial founder. Does this mean the person has built and sold something? Have they been able to complete successful raises? Or is it possibly they just have created multiple ideas?</p><p>I imagine that just like women and minorities don&#39;t apply for jobs unless they are 100% qualified, many don&#39;t call themselves founders until they have reached some invisible, self-governed milestone. It takes a lot of audacity to call oneself a founder, and I, and the founders I work with, are not often rewarded for audacity.</p><p>So yes, I&#39;m a founder. I&#39;m a founder because I say so. I&#39;m a founder, and then I can add to the conversation to tell people what I&#39;ve accomplished. There was no metric or benchmark I needed to pass to get there.</p><p>Let&#39;s all have that audacity. Let&#39;s all get out there and dream.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=d85b2c998a93" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Hey UX, AI does not change the game]]></title>
            <link>https://medium.com/design-bootcamp/hey-ux-ai-does-not-change-the-game-af8fdf1946d5?source=rss-6d4e528f05b2------2</link>
            <guid isPermaLink="false">https://medium.com/p/af8fdf1946d5</guid>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[technology]]></category>
            <category><![CDATA[ux]]></category>
            <category><![CDATA[artificial-intelligence]]></category>
            <category><![CDATA[user-experience]]></category>
            <dc:creator><![CDATA[Clare Cady]]></dc:creator>
            <pubDate>Tue, 25 Nov 2025 16:02:34 GMT</pubDate>
            <atom:updated>2025-11-25T17:56:57.662Z</atom:updated>
            <content:encoded><![CDATA[<p>At least, not as much as you think it does.</p><figure><img alt="Image of a black background with the word change in white font in all capital letters. The same word is reflected upside down at the bottom of the image, with an X drawn through it so that it looks like the word has been broken like glass." src="https://cdn-images-1.medium.com/max/1024/1*MC-iObYdYnNCl0DLDDGsDw.jpeg" /></figure><p>I had the privilege of being hosted by the Boston chapter of IxDF recently, to give a talk entitled, “They call it HCI for a reason: finding the balance between humans and AI in UX.” It was a fabulous evening speaking to a very engaged, sold out audience. And when I say engaged… I mean fully in it with opinions, ideas, and questions. Never have I had such a wonderful and robust Q&amp;A session in my over 10 years of public speaking!</p><p>A theme emerged as we entered into the discussion portion of the evening. I found myself fielding questions about the future of UX as we both build and use generative AI tools. A lot of people were asking how these tools will change the game, and how to approach design and research specifically when building AI tools. The theme that emerged is this: a lot of UX professionals seem to completely forget that we have figured a lot of this out already, and that AI doesn&#39;t really change the game all that much.</p><p>Here&#39;s an example.</p><p>One of the questions I was asked was how does AI change the UX research process? I would argue that despite there being some new ways to test tools, because the tools themselves are different, that the process of conducting user research does not change just because you are building AI products. You still need to start with the problem statement, determine what users are impacted by the problem, engage with those users to understand and discover, validate designs, iterate along the way, communicate well with engineers, and assess impact after you have shipped the product.</p><p>This hasn&#39;t just come up at my public talks. In a recent conversation, I was chatting with a designer about ways that we can mitigate risk and address potential harm to users when using generative AI in user workflows. The designer was concerned that “unleashing a chatbot” on the users in the workflow could lead to significant problems, regardless of how carefully we trained it. Upon digging more deeply into the product requirements it became very clear that we didn&#39;t need to let a user run amok in an LLM. Instead, we could apply the basic UX heuristic of constraints and allow the user to select from very specifically tested prompts. The user interface could serve as a buffer between the user and the LLM, and protect them by guiding the workflow.</p><p>Sometimes it seems like when it comes to AI, we toss the baby out with the proverbial bathwater. Like the introduction of this specific type of tool negates decades of tried and true processes and methods. I can see where this might be coming from. The broad brush messaging we get about AI is that it is changing the game entirely. It&#39;s disrupting literally everything. It&#39;s going to transform our jobs and work and products and the world in ways we haven&#39;t even imagined… And it&#39;s going to do it at lighting speed. It can be easy to get swept up in this type of rhetoric and not take a pause and realize that AI is just a new place to apply our old craft.</p><p>And by &quot;old,&quot; I mean old in tech years, not human years. I apply the equation that one human year equals five tech years.</p><p>The profession needs to slow down and take a breath. We need to center ourselves on the things that we know work and consider how they will work, even in this new context. We need to remember that we are experts, and that we have the tools we need to be successful even in the face of change. We also need to realize the difference between perceived and actual change. I think there will be some really amazing innovations in how we do our work, and I also think a lot of it will remain the same.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=af8fdf1946d5" width="1" height="1" alt=""><hr><p><a href="https://medium.com/design-bootcamp/hey-ux-ai-does-not-change-the-game-af8fdf1946d5">Hey UX, AI does not change the game</a> was originally published in <a href="https://medium.com/design-bootcamp">Bootcamp</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Gen AI is not the only source of slop]]></title>
            <link>https://medium.com/@clarelcady/gen-ai-is-not-the-only-source-of-slop-67af2f1b79df?source=rss-6d4e528f05b2------2</link>
            <guid isPermaLink="false">https://medium.com/p/67af2f1b79df</guid>
            <category><![CDATA[ai-slop]]></category>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[generative-ai]]></category>
            <category><![CDATA[writing]]></category>
            <category><![CDATA[ai-ethics]]></category>
            <dc:creator><![CDATA[Clare Cady]]></dc:creator>
            <pubDate>Mon, 24 Nov 2025 14:05:42 GMT</pubDate>
            <atom:updated>2025-11-24T14:05:42.636Z</atom:updated>
            <content:encoded><![CDATA[<p>Sometimes it’s fully human-made, and just as useless.</p><figure><img alt="The back end of a black van is at the top of an image. The back tire is partially buried in wet, thick, red clay mud. There are tire tracks all around, implying that many vehicles have been through this area." src="https://cdn-images-1.medium.com/max/1024/1*T0dqnMQjwgFm-f3xLUHxaQ.jpeg" /></figure><h3>People are sloppy too</h3><p>There’s one type of AI slop we aren’t really talking about. It’s not the crappy videos of Trump playing soccer in the oval office, or clips of rabbits jumping on a trampoline, or the emoji-bulleted, soulless, copy/paste jobs all over LinkedIn. It’s something a little more subtle, but is equally a big waste of my time on text-based scrolling platforms.</p><p>I’m talking about AI “opinion” pieces that are devoid of actual opinions.</p><p>Here are some examples (without naming names):</p><ul><li>Tutorials on how to tell if someone is writing with AI.</li><li>Reposts of AI articles without any analysis.</li><li>“Hot takes” on AI that are designed as rage or click bait rather than discourse.</li><li>Late-to-the-game opinions that read like regurgitation of what other people have been saying for months.</li></ul><p>Stop it people — please?</p><p>I imagine if I were to go back through my LinkedIn posts I could find examples of where I too have been guilty of this. The FOMO speed-to-market pace of AI tools we saw earlier this year often felt like it necessitated lightning speed content creation to just keep up and stay relevant. So please, if you want to call me out, that’s fine. I’ve beaten you to the punch.</p><h3>Why this is a problem</h3><p>We are living inside a significant cultural shift — history in the making. I both love and hate this for us. I love that we have a chance to be a part of it and to shape it in a way that could, if done well, make our lives better. I hate that there is just so much crap out there that we are going to miss the signal through the nose. There’s often too much going on for us to find a discourse, shape real ideals, and think about what we want the future to look like as AI tools improve and become more prevalent. That’s the space I want to be in, and I want others to join me there. Otherwise we lose control of the narrative and can’t continue to critique, iterate, protect, and develop a future that has real potential.</p><h3>So what do we do?</h3><p>If you want to create content about AI, please keep doing it. All I ask is this:</p><p>Have a perspective.</p><p>Pause, think, and develop a real opinion. Share it. Engage with others around it. Let it be fluid — mutable — as you learn more and mature your views. Don’t troll or argue. Pose questions. Be curious. Advocate for others through critique and perhaps even (gasp) read some perspectives that don’t align with your own. Stay current. Be constructive. Tag people whose perspectives you value, and get their opinion on it. Figure out what it means to not only exist inside of history, but to be a part of shaping it.</p><p>Yes, this will take longer. You may have to spend some time deciding what your perspective really is. You may not be able to use generative AI to create the content. You may have to write it yourself. You may have to work on some new prompts to help you make sense of what’s in your brain. It will take more work, but the work will be worth it. I promise.</p><p>Let’s stop creating slop we need to wade through, and build a flow that actually helps us get where we want to go.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=67af2f1b79df" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Will we stop solving for the AI “problem?”]]></title>
            <link>https://medium.com/design-bootcamp/will-we-stop-solving-for-the-ai-problem-742865d52443?source=rss-6d4e528f05b2------2</link>
            <guid isPermaLink="false">https://medium.com/p/742865d52443</guid>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[artificial-intelligence]]></category>
            <category><![CDATA[technology]]></category>
            <dc:creator><![CDATA[Clare Cady]]></dc:creator>
            <pubDate>Wed, 12 Nov 2025 13:31:20 GMT</pubDate>
            <atom:updated>2025-11-13T12:14:20.127Z</atom:updated>
            <content:encoded><![CDATA[<p>Can we all agree it’s not magical fairy dust?</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*2sILIWvDTtzuOKKme3UuuA@2x.jpeg" /><figcaption>Photo by <a href="https://firmbee.com/freebees">Firmbee.com</a> on <a href="https://unsplash.com/?utm_source=medium&amp;utm_medium=referral">Unsplash</a></figcaption></figure><p>I’ve written before about how the hype around AI seems to be cooling off, and we are entering into what I’ve dubbed AI fall. As a person who follows the seasons and tries to take some cues from nature, fall as a time to reap the benefits of your hard work of the summer before settling in for a winter of benefitting from the abundance that you have created for yourself.</p><p>The problem is when it comes to AI there are so many who are not going to reap or benefit. Even the biggest companies are operating at massive losses. Smaller companies are folding, or frantically trying to do additional raises. And we now know that 95% of AI initiatives in business have failed. What was assumed to be a massive year for profitability and productivity is now being seen as the bursting of another tech bubble.</p><p>Does this mean that AI has failed? I would argue no, and that there are a lot of places where there have been successes. I also think that the best of AI is yet to come.</p><p>There are both large and small companies out there taking their time to bring AI tools to market, and I believe that when they do, these will be smarter, have a better user experience, and be more effectively focused on solving real world problems. I’ve had the good fortune to work with the number of founders that I would place into this bucket, and I’m telling you right now, when they get out there, they will be making people’s lives better.</p><p>One of the things I hope we do as an industry is to learn from some of the things we’ve seen that didn’t work. From where I sit, one of the biggest issues was panic development of AI powered products across industries. Entrepreneurs and founders went out there thinking “how can I bring AI to this market?” The focus was on AI as a capability and how it wa<em>s obvious</em>ly going to bring value to everything.</p><p>That’s just not the case, and it’s not how great products are built. Great products are built by solving problems.</p><p>That’s this jumped the proverbial rails. If the problem you think you’re solving for is that AI is not in a specific market or industry, you haven’t validated whether or not people actually want or need to use it. You haven’t taken a look to see whether or not AI is actually the best solution for a problem, if you have even identified one. This is where the products that were slower to come to market seem to have done a better job. Companies took a look at the actual problems. Their users had, and determined that AI was the best way to either solve them, or improve the solution.</p><p>My LinkedIn feed was packed this year with smaller companies, touting AI and making bold claims that they were going to take down massive companies like Figma. Then at Config the Figma PMs marched out one by one announcing a suite of AI powered tools that we’re going to augment what they already offer. The bold claims in my feed changed from how Figma was going to be taken down, to how it had killed a number of smaller startups.</p><p>Now mind you, I understand that a company like Figma has the luxury of being able to invest in AI tools over time, as well as the competitive advantage of the fact that designers are already using their tools on a daily basis. I’m not saying that smaller companies shouldn’t try and compete with them. I imagine that the tools that Figma has released and are refining are going to be better because there are smaller companies nipping at their heels. I also don’t think they are all necessary, and some will probably be scrapped. What’s good here is we don’t want the 500 pound gorillas in the room to think no one is going to challenge where they sleep, and those smaller companies were doing just that.</p><p>But when you choose to go to market simply infusing AI into something that already exists, you’d better be sure that AI is a truly important differentiator. You need to make sure that AI is necessary for the tool you are trying to build. And you need to make sure that people want it. I don’t think a lot of founders, did either of these things. When you look at a public perception of AI, it is clear that not everybody is interested in having it in the things they use. When it comes to tech, there is a love-hate relationship with it. It presents tremendous potential, while also threatening people and their livelihoods. When you look at the quickly declining adoption rates for tools like Lovable or V0, it’s not hard to wonder if the critical differentiator isn’t AI but something else. Or even that there isn’t a critical differentiator at all.</p><p>I do wonder what would’ve happened if the founders that took the pathway of AI-industry-solution instead took a pathway of problem-solution-AI. What would’ve happened if they had honed in on the problems that users had with some of these existing products and experiences, thought through the solution, and determined that AI was the best thing for the job?</p><p>I think a lot of companies would have decided AI was not in fact the way to go.</p><p>This means that there would’ve been fewer products on the market. It would mean that there are fewer founders that got massive investments and valuations. I’m not going to pretend that those things don’t drive a lot of the market. It’s not necessarily about innovation, but about looking good and making a lot of money. The realist in me understand that those things are greater incentives than following effective process, and I know that we all have different definitions of what “effective” is in the first place.</p><p>However, I do think there are many things that would have been significantly better in tech if there had just been a pause and a determination prior to pushing straight into AI. You just need to take a look at the jobs market to see the damage. Again, referencing my LinkedIn feed, I see the pain and frustration of this daily. At this point, the layoffs aren’t just because people are being seen as expendable because AI products are going to fill the gap. It’s also because companies have gone in so hard on AI that they need to cut ranks as their investments do not return the proper dividends.</p><p>So what can we learn from this?</p><p>Companies large and small need to focus on user and problem-centric solutions. This may mean that AI is not a part of the equation. It <em>also</em> may mean that AI is part of the equation, and that using it as part of the solution is truly a differentiator. We’ve already seen this bringing us some measure of success, and I think that those who have been doing this are going to bring truly innovative and effective products market.</p><p>We can’t forget to go upstream before we hit the rapids. My hope is that as we enter AI winter that is a lesson we learn.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=742865d52443" width="1" height="1" alt=""><hr><p><a href="https://medium.com/design-bootcamp/will-we-stop-solving-for-the-ai-problem-742865d52443">Will we stop solving for the AI “problem?”</a> was originally published in <a href="https://medium.com/design-bootcamp">Bootcamp</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Your “fast paced” and “high commitment” startup isn’t cool and edgy — it’s gatekeeping]]></title>
            <link>https://medium.com/@clarelcady/your-fast-paced-and-high-commitment-startup-isnt-cool-and-edgy-it-s-gatekeeping-e90675457422?source=rss-6d4e528f05b2------2</link>
            <guid isPermaLink="false">https://medium.com/p/e90675457422</guid>
            <category><![CDATA[gatekeeping]]></category>
            <category><![CDATA[workplace-culture]]></category>
            <category><![CDATA[startupş]]></category>
            <category><![CDATA[team-building]]></category>
            <category><![CDATA[recruitment]]></category>
            <dc:creator><![CDATA[Clare Cady]]></dc:creator>
            <pubDate>Thu, 30 Oct 2025 18:39:11 GMT</pubDate>
            <atom:updated>2025-10-30T18:39:11.931Z</atom:updated>
            <content:encoded><![CDATA[<h3>Your “fast paced” and “high commitment” startup isn’t cool and edgy — it’s gatekeeping</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*sJ_LqATIR7Vto-GclBqeiQ.jpeg" /></figure><p>And you are going to miss out on great talent</p><p>If you are a founder recruiting with this kind of language and expectations, this post is for you:</p><ul><li>Be ready to work long hours… sometimes as much as 80 hours a week.</li><li>We work impossibly fast and you will need to keep up.</li><li>This job is not for the passive. You must be relentless and obsessed.</li><li>You need to eat, sleep, and breathe the product.</li><li>Late nights and weekends are expected.</li></ul><p>I <em>want</em> to understand why you are doing this. Maybe you want to sound cool and edgy. Maybe it simply reflects the culture of your company (and if it does at least you are up front about it). Maybe you are trying to cut through the noise of the unemployed masses in tech right now with a signal that is going to thin the onslaught of applications. Maybe you just don’t know what the heck you are doing, but this sounds good to you.</p><p>I am going to assume the best here, and say it’s one of those things. I am going to assume you are interested in just getting a team up and running fast so you can ship cool things. I am going to assume you are so focused on not being left behind that you are overlooking how these kinds of expectations can hurt others, and in the long run, you.</p><h4>Who you leave behind</h4><p>When you set these kinds of intense job requirements you may find awesome people who can meet them. You will also throw up massive, impossible barriers to people who really, really can’t. There are so many kinds of people who are not able to meet these kinds of expectations. Here are some examples:</p><ul><li>Single parents</li><li>Women who work the proverbial third shift</li><li>People who can’t afford to live near your office</li><li>People with disabilities and chronic illness</li><li>People who are older</li></ul><p>When you post those kinds of requirements, you send a clear message to these people “do not apply” because the workplace you are curating is immediately and undeniably toxic to them. Not everyone can live and breathe your company. Not everyone can be a solitary worker with no one relying on them beyond the workplace. Not everyone can physically meet those demands. Not everyone can afford to live near or commute to your expensive office. FWIW, I think these kinds of expectations are toxic to anyone, but that’s not the point of this article.</p><p>PS: if my assumptions are wrong, and you do know what you are doing… shame on you.</p><h4>Why this is bad for the bottom line</h4><p>If you curate a team of people who have zero commitments, constraints, or boundaries, you will probably get decent to good results… for a while. But there are serious consequences that can occur when this is the culture you start up with. Employees will most definitely burn out. They will leave with so much institutional knowledge <em>in their heads</em> because you are probably also moving so fast you are leaving out documentation.</p><p>In addition, you have created an army of obsessed sycophants who are just. like. you. While this might seem great at the onset, when you lack a range of perspectives, experiences, and experience levels on your team you can develop a product echo chamber. Sure, you may be in lock step with each other, but you are all going to miss the same things in pursuit of the same thing. Even if you are smart enough to loop in your customers, you are still going to view their thoughts and feedback from the same lens. Innovation is stifled when everyone is the same.</p><p>AND the types of people I mentioned are left out (and I have only named a few) not only will bring in that range of perspectives, but also, they are often times older with more experience in their fields. Why? Seriously, <em>why,</em> would you leave that on the table? In a time when we are dealing with rampant ageism in tech <em>you</em> could be the founder who is reaping the benefits of that massively talented pool of professionals. These people will be more efficient, and deliver higher quality results in less time. And NO — they are not unable to adapt to AI tools. They are a generation of people who grew up adapting to new tools. They are probably <em>better</em><strong><em> </em></strong>at it than you.</p><p>In short, your approach is myopic, exclusive, and will hurt you in the long run.</p><h4>What do you do?</h4><p>In an industry of risk, I find it laughable that I need to call this risky. But maybe that’s what it is, so I will ask a few questions. Because if you as a founder were to pause, think about what you <em>really </em>need for your success, and go from there, I think you yourself might be happier. Do you really want to model 80 hour work weeks? Do you really think that running at that pace actually equates to more success? Are you bought into the intoxicating hype or victim to the FOMO? You are hurt by toxic culture as well — why are you perpetuating it? Why would you pass on that talent? Are you meaning to gatekeep? Were you thinking about what it means to start up this way? What happens in the future? Is this sustainable?</p><p>Only you can answer those questions truthfully.</p><p>And PS again: if you know what you are doing here… shame on you.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=e90675457422" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Will we need UX data engineers?]]></title>
            <link>https://medium.com/design-bootcamp/will-we-need-ux-data-engineers-0137c598f427?source=rss-6d4e528f05b2------2</link>
            <guid isPermaLink="false">https://medium.com/p/0137c598f427</guid>
            <category><![CDATA[future-of-work]]></category>
            <category><![CDATA[artificial-intelligence]]></category>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[ux]]></category>
            <category><![CDATA[llm]]></category>
            <dc:creator><![CDATA[Clare Cady]]></dc:creator>
            <pubDate>Tue, 28 Oct 2025 13:39:00 GMT</pubDate>
            <atom:updated>2025-10-28T14:10:41.610Z</atom:updated>
            <content:encoded><![CDATA[<p>LLMs are part of the user experience — but how do we design them well?</p><figure><img alt="A close up of a computer screen where white and yellow code is typed on a black background." src="https://cdn-images-1.medium.com/max/1024/1*2vL353I9PE6yB1lhIdSbfw.jpeg" /></figure><p>I was speaking recently about the balance between humans and AI and user research and design. It was a great topic, a great venue, a great crowd, and a great conversation. One of the things that came up was the role of UX professionals in ensuring that AI products deliver quality experiences to users. I was asked if I felt that the LLM is part of UX.</p><p>I most certainly do.</p><p>When humans interact with software, we develop and design them for minimal friction, optimal efficiency, and maximum delight. Our goal is to solve users&#39; problems while leaving them better off than if they had solved the problem themselves, and helping them feel a positive emotion when the workflow is complete. Our goal is to make people&#39;s lives better.</p><p>This goes for generative AI as well. What&#39;s interesting here is that the workflows we use to interact with gen AI tools are relatively simple. Most chatbots are only a few elements: a place to enter text, a button to send, and an interface to read the messaging. If it is more sophisticated than that, it may allow for people to speak instead of type, and then there are going to possibly be places where you can see saved conversations. It is highly simplistic because the complexity is not in the workflow, but instead in the outputs of the LLM.</p><p>So then, the user experience is transferred to whatever the LLM responds with when a user prompts it.</p><p>What does this mean for UX? To start, it means that we need to thoughtfully adhere to simple workflows that won&#39;t make it harder for a user to interact with the LLM. This is not a place for us to be overly innovative or create more complexity in the system than is necessary. After that, there is a need for critical and thorough user research to ensure that the outputs of the LLM are effective, usable, and trustworthy. There is a need to ensure there is proper tone to match the expectations of the user based on the context in which both the user and the chatbot exist. There also needs to be a way to get out of situations that are problematic, correct the LLM when necessary, and rebound when something goes wrong.</p><p>One of the problems I brought up during our discussion is that so many of the generative AI tools out there are a user interface over the top of an API call to another system. This leaves the bulk of the fine-tuning of the LLM outputs to prompt engineering, and the entire system at the mercy of changes made to the source from which the API call draws. I can only imagine the havoc that the change from GPT 4 to 5 may have wreaked on tools built on top of it.</p><p>The best way to address this issue is to utilize private data sets instead of API calls to third parties. This can certainly be a challenge because many companies do not have access to the proper data sets to do their work. However, if they are able to do this, they will have both increased control over how the models are structured, and also can refine outputs more easily. A narrowly drawn LLM trained on a data set focused on nutrition for example, is less likely to pull incorrect information from the internet and can provide sounder and more useful advice to the user. While it may be more work upfront, there is less likelihood for misinformation, bias, and drift when the data set is known and tested against the user persona and their most relevant use cases.</p><p>So this begs the question: will the rise of LLMs lead to the need of an additional layer of UX expertise? I&#39;m thinking something like a UX data engineer that is focused on developing data sets that will most effectively meet user needs. While I know many companies do have really excellent data engineers, I do wonder whether or not they are training on things like personas, UX heuristics, and the critical elements of the user experience.</p><p>As we continue to see a shake-up in the field, with the flattening of roles and the addition of AI expertise as part of job descriptions, will we also see the inclusion of data and data science as user experience into the equation? I think this will be both an interesting and exciting area to watch.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=0137c598f427" width="1" height="1" alt=""><hr><p><a href="https://medium.com/design-bootcamp/will-we-need-ux-data-engineers-0137c598f427">Will we need UX data engineers?</a> was originally published in <a href="https://medium.com/design-bootcamp">Bootcamp</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Are we looking at a revival of the UX generalist?]]></title>
            <link>https://medium.com/design-bootcamp/are-we-looking-at-a-revival-of-the-ux-generalist-e841d5fbf5b1?source=rss-6d4e528f05b2------2</link>
            <guid isPermaLink="false">https://medium.com/p/e841d5fbf5b1</guid>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[job-market]]></category>
            <category><![CDATA[ux]]></category>
            <category><![CDATA[generalists]]></category>
            <category><![CDATA[artificial-intelligence]]></category>
            <dc:creator><![CDATA[Clare Cady]]></dc:creator>
            <pubDate>Thu, 23 Oct 2025 16:50:50 GMT</pubDate>
            <atom:updated>2025-10-23T18:25:59.534Z</atom:updated>
            <content:encoded><![CDATA[<p>UX JDs seem to suggest it’s happening.</p><figure><img alt="A group of people lean over a white table with pens and sticky notes in their hands. On the table is an open laptop, sticky notes with writing on them, and wireframe diagrams drawn on white paper." src="https://cdn-images-1.medium.com/max/1024/1*jVsdplY6eb5FJ2RijJz7UQ.jpeg" /></figure><p>Unless you are hiding under a rock, you are aware that UX roles in tech are in decline. A <a href="https://blog.academyux.com/ux-job-market-analysis-q1-2025/?utm_source=chatgpt.com">recent jobs analysis</a> indicated that the number of listed roles for UX professionals dropped below 1000 in Q1 of 2025. And while we are now in the midst of the mythical “September surge,” where companies make their final hires when the budget gets locked for the year, it does not appear that the number of open roles will return to those of previous years. UX job growth has dropped by 38% this year per the <a href="https://www.weforum.org/publications/the-future-of-jobs-report-2025/in-full/2-jobs-outlook/">World Economic Forum.</a> The fears that many UX professionals have been voicing this year about UX in the age of AI are coming true, and it leaves me (and probably others) wondering what comes next.</p><p>What I am seeing is an emerging trend that when UX roles are being posted, they are more senior and with the <a href="https://www.nngroup.com/articles/return-ux-generalist/?utm_source=chatgpt.com">expectation of doing more than design.</a> UX and product designers are being asked to take on research tasks, as are UX Strategists. In addition to that, design and research are often showing up not in UX roles, but instead as part of the job description of a Product Manager. It appears that as AI tools are able to make UX and product professionals more efficient, companies are deciding which skills and functions are most important — design and product management — and taking the skills they seem less important — research and writing — and passing those duties on to the designers and PMs. There is also (in my own observations) an increase in UX strategist roles, which suggests that process is also important, and that research can be done by anyone.</p><p>I should note that there is no mention of UX writing in most of these JDs. It’s research only that is being added as a responsibility. So writing? Clearly the lowest priority.</p><p>It’s not new for there to be UX generalists in the world of tech. There was a time when UX research was not scaled up as a major discipline with an increase in specialization over the past 5–10 years. Some of this could be because companies were moving too fast and doing too much for UX professionals to do more than just design, so research was split off as a new role. It could be that companies realized that research rigor was best left to, well, researchers who understood how to design powerful studies. And as the field improved things like onboarding and recognized the importance of language in products, good writing also became more prized, leading to more UX writer roles.</p><p>But as AI makes it easier and faster to do all of these UX tasks, companies are starting to make moves to flatten the roles. From where I sit, there seem to be three different approaches that tend to align with the level of UX maturity of the company.</p><p><strong>Low maturity: </strong>these companies are fully flattening roles. They may even put design and research onto the PM role instead of hiring any UX professionals, relying instead on vibe coding products to create prototypes based on the PMs own research.</p><p><strong>Mid-level maturity:</strong> these companies are hiring UX generalists and relying on them to leverage AI to be effective in research, design, and writing instead of hiring specialists.</p><p><strong>High maturity:</strong> these companies are still hiring specialists for the most part, though some have also pivoted into the more flattened roles.</p><p>What does this mean for the field? First and foremost, it means that there are going to be fewer roles… at least for now. Companies are still trying to figure out the right combination of personnel, tooling, and financial strategy to be cash flow positive and avoid racking up UX debt in their products. This will go beyond the number of roles that come available, and will delve into the types of roles and what they will be expected to do. We will most certainly see an ongoing movement toward generalist roles as both AI tools improve, and aspractitioners get better at using them. We may see a transformation in roles like research or writing, or simply fewer of them as well. These disciplines may be absorbed in some cases, or there may come a reckoning across tech — something that indicates that it’s wiser to invest in specialized work. PMs may be taking on more responsibilities as product tools become more effective. While it’s nearly impossible to say what the changes <em>will</em> be, it is clear that change is already here, and it’s likely to speed up before it stabilizes.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=e841d5fbf5b1" width="1" height="1" alt=""><hr><p><a href="https://medium.com/design-bootcamp/are-we-looking-at-a-revival-of-the-ux-generalist-e841d5fbf5b1">Are we looking at a revival of the UX generalist?</a> was originally published in <a href="https://medium.com/design-bootcamp">Bootcamp</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[The ethics of UX research in AI: questions the field needs to answer]]></title>
            <link>https://medium.com/design-bootcamp/a-the-ethics-of-ux-research-in-ai-questions-the-field-needs-to-answer-f2782e1d21cc?source=rss-6d4e528f05b2------2</link>
            <guid isPermaLink="false">https://medium.com/p/f2782e1d21cc</guid>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[ux-research]]></category>
            <category><![CDATA[artificial-intelligence]]></category>
            <category><![CDATA[ux]]></category>
            <dc:creator><![CDATA[Clare Cady]]></dc:creator>
            <pubDate>Mon, 06 Oct 2025 16:35:09 GMT</pubDate>
            <atom:updated>2025-10-06T21:02:52.934Z</atom:updated>
            <content:encoded><![CDATA[<p>I have so many questions</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*_84iw57m8M4Wkb-w5MsEgg.jpeg" /></figure><p>I held space for an unconference session at <a href="https://www.uxyall.org/">UX y&#39;al</a>l just recently. My topic was AI ethics in UX research. If you don&#39;t know what an unconference is, here is an <a href="https://www.forbes.com/sites/rebeccabagley/2014/08/18/how-unconferences-unleash-innovative-ideas/">article</a> that explains it pretty well. Essentially, participants drive the conversation or present on topics that their own choosing. For UX Y’all, there was only one session that offered an unconference format. We were able to submit a topic suggestion and mine was selected for the session.</p><p>One of my favorite things about the session was that I didn&#39;t actually lead it. Instead, a group of amazing UX researchers joined me at the table with only the topic &quot;AI ethics in UX research&quot; as a prompt. I had some preconceived notions about what our conversation would be. I thought we might talk about the use of note takers, data security in LLMs, or whether or not synthetic users are an ethical way to conduct research. But that wasn&#39;t the case at all.</p><p>Instead, we launched into a deep conversation about what it means to conduct the research that supports the creation of AI products that eventually replace (or could replace) human jobs. In the 45 minutes we had to discuss, we were really only able to reflect on and form some critical questions that the field will need to grapple with as we move forward in research on AI-powered products. Here are a few of the higher-level ones that we touched on.</p><ul><li>If we are conducting research on a product, that &quot;makes people more efficient,” and that product might lead to those “more efficient” people losing their jobs, <strong>whose problems are we actually solving?</strong></li><li>If we aren&#39;t solving the users’ problems, <strong>are we even conducting UX research?</strong></li><li>If the purpose of UX is to make people&#39;s lives better, <strong>what are the ethical implications of conducting research on efficiencies that don&#39;t?</strong></li><li><strong>What does it mean to advocate for the user</strong> if both the user and the researcher are concerned that the suggested changes will lead to job loss?</li><li>If UX research is meant to help identify what could go wrong,<strong> where is the line between issues with the product, and issues with how the product is leveraged?</strong> If what could go wrong is that people could lose their jobs, should we be flagging that?</li><li>What is the role of the researcher in ensuring there are humans in the loop? <strong>Is this our responsibility?</strong></li><li>Since we are often the front line for new products with users, <strong>what is our role in socializing AI as part of the future of work?</strong></li></ul><p>And finally,</p><ul><li><strong>Is helping build these types of efficiencies, knowing what we know, even ethical?</strong></li></ul><p>While we did chat around these things and share examples of what we are seeing in our work, these questions are so large and so complex that we certainly didn&#39;t solve them in the 45 minutes we had together. What I feel we did do is define some very important conversations that we must be having in the field. And while this conversation was had by researchers with the context of UX research, I would argue that these questions certainly expand into all of UX, and could also be applied to product. I mean, I could say that they just go out to everyone… but at that point I&#39;ve become so abstract it would be hard to do anything with them.</p><p>The jaded part of me wonders if this is one of the reasons that UX professionals are finding themselves cut out of the loop when it comes to AI. Even when we aren&#39;t focused on AI-powered products, we often find ourselves stuck in a battle between business goals and user needs. As this conflict intensifies, can we assume that one way to avoid having to grapple with some of these ethical questions is to simply take UX practitioners out of the picture? Mind you, I&#39;m not saying that the other people we work with in tech aren&#39;t willing to have ethical conversations. Instead, what I&#39;m highlighting is the fact that the people whose job it is to hold that space are being removed from the room. Is the UX debt that I&#39;ve been writing about also indicative of a larger ethical debt?</p><p>I&#39;d like to explore these things in the coming months. Stay tuned.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=f2782e1d21cc" width="1" height="1" alt=""><hr><p><a href="https://medium.com/design-bootcamp/a-the-ethics-of-ux-research-in-ai-questions-the-field-needs-to-answer-f2782e1d21cc">The ethics of UX research in AI: questions the field needs to answer</a> was originally published in <a href="https://medium.com/design-bootcamp">Bootcamp</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[UX Researchers are AI ambassadors]]></title>
            <link>https://medium.com/design-bootcamp/ux-researchers-as-ai-ambassador-8cd0cac5df73?source=rss-6d4e528f05b2------2</link>
            <guid isPermaLink="false">https://medium.com/p/8cd0cac5df73</guid>
            <category><![CDATA[ux-research]]></category>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[artificial-intelligence]]></category>
            <category><![CDATA[ai-ethics]]></category>
            <category><![CDATA[user-research]]></category>
            <dc:creator><![CDATA[Clare Cady]]></dc:creator>
            <pubDate>Tue, 23 Sep 2025 15:10:27 GMT</pubDate>
            <atom:updated>2025-09-23T17:24:08.120Z</atom:updated>
            <content:encoded><![CDATA[<p>We are often the front line in trust</p><figure><img alt="An apple laptop sits open on a desk with the side facing towards you. Two people are working at the laptop with their hands the only thing showing. Two hands, one from each person, are pointing at the screen, and one of the users is moving their hand on the touch pad." src="https://cdn-images-1.medium.com/max/1024/1*XU8-_DwhedfBiPfawE7UKA.jpeg" /></figure><p>I&#39;ve been interviewing for roles in UX research now for three months. In most cases, the companies I am chatting with are building AI powered products. This has been really exciting for me as it is an area I really enjoy. Talking about the intersection of AI and UX research has led me to do some reflection on what it means to be a UX researcher and how it might change as we continue to interface with users on AI powered products.</p><p>Most of my experience has been in B2B SaaS products, which means that while there are plenty of people, including myself, who talk to the buyers of the products, fewer people are talking directly to the users who are in the workflows on a day-to-day basis. Mind you, I&#39;m not saying that customers can&#39;t be users, but it is not uncommon for those who are purchasing to not be using the product regularly. This can cause some disconnection between what a buyer is trying to accomplish and the user experience on the other end. I cannot tell you how many times I have been talking to a user who is telling me something that is in direct conflict with feedback we have gotten from the buyers at the same company. In these cases, I find that my role starts to become not only data collector and insights deliverer, but also diplomatic bearer of incongruent news.</p><p>I don&#39;t mind this role. I worked for a very long time in conflict and crisis management, so in many ways I am uniquely suited to handle it.</p><p>As more and more AI powered tools get into the market, buyers are chasing the value they promise to add. <a href="https://medium.com/design-bootcamp/ai-has-a-ux-debt-problem-ef9980e2d9ad">I&#39;ve written before</a> that a lot of these tools are going to market too quickly and without the proper UX research and design that will make them intuitive and delightful to use. The gap between the promise of these tools and the experience of them seems to be even wider than with other tools I have researched. In addition, when AI is introduced into the equation, there are additional considerations around trust and safety that must be addressed at the user level.</p><p>This means that UX researchers are now not only facing the front line of a potential gap between what customers and users think about a product, but also what they think about AI. We are ambassadors for the product and for the technology itself, having to manage expectations on how the tools are being positioned, and how the company is demonstrating and expressing its commitment to transparency and safety. While these things may be handled with buyers, there is a high likelihood that users will not have been exposed to the responsible AI conversation at all.</p><p>What this means for researchers is that they should not only be versed in personas, workflows, road, maps, and other aspects of product development, but they should also be well aware of how their company is approaching AI as a tool. This means they should be able to point to any public facing transparency statements, governance statements, or other policies around AI and it&#39;s use in the products. Not being aware of these things could possibly cause further concern or distrust in a user because their primary contact or window into the product cannot help assure them of the safety of their data or their experience using the tools.</p><p>It feels like a heavy responsibility. However, this is just another layer on top of the ways in which user researchers already serve as ambassadors for their companies. Someone who already takes that role very seriously has likely already done the work to cover these additional user needs.</p><p>If that&#39;s you, well done! If it&#39;s not, I highly recommend you do the work to get there. You will be a much stronger researcher for it.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=8cd0cac5df73" width="1" height="1" alt=""><hr><p><a href="https://medium.com/design-bootcamp/ux-researchers-as-ai-ambassador-8cd0cac5df73">UX Researchers are AI ambassadors</a> was originally published in <a href="https://medium.com/design-bootcamp">Bootcamp</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
    </channel>
</rss>