<?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 Delethia Johnson on Medium]]></title>
        <description><![CDATA[Stories by Delethia Johnson on Medium]]></description>
        <link>https://medium.com/@delethia?source=rss-cb64445e7d1a------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*nayDiWt2O079ailXHiNjHg.webp</url>
            <title>Stories by Delethia Johnson on Medium</title>
            <link>https://medium.com/@delethia?source=rss-cb64445e7d1a------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Sun, 22 Nov 2026 08:42:17 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@delethia/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[When the Future Couldn’t See Me: Understanding AI Bias]]></title>
            <link>https://medium.com/@delethia/when-the-future-couldnt-see-me-understanding-ai-bias-3ffdc37b6dc0?source=rss-cb64445e7d1a------2</link>
            <guid isPermaLink="false">https://medium.com/p/3ffdc37b6dc0</guid>
            <category><![CDATA[technology]]></category>
            <category><![CDATA[ai-bias]]></category>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[ethics]]></category>
            <category><![CDATA[data-science]]></category>
            <dc:creator><![CDATA[Delethia Johnson]]></dc:creator>
            <pubDate>Thu, 22 Jan 2026 17:11:12 GMT</pubDate>
            <atom:updated>2026-01-22T17:11:12.529Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="An African American woman with a dark complexion looking at a futuristic face detection machine and the machine did not recongnize her face." src="https://cdn-images-1.medium.com/max/1024/1*WKgYwNgfUwmwdzI6Z_kW6A.jpeg" /><figcaption>AI generated recreation of my experience during a museum visit in 2023</figcaption></figure><p>The room buzzed with excitement: voices in different languages, the click of phone cameras, announcements echoing over loudspeakers. I was standing in line on the “Tomorrow, Today” floor with my group, waiting for our turn to try an AI facial recognition exhibit. The display was meant to showcase AI’s role in public safety.</p><p>One by one, people stepped up to the machine. Faces were recognized. Labels appeared. The technology seemed to work perfectly until it was my turn.</p><p>I walked up to the camera and waited. Nothing happened. Thinking the machine had broken, I waved my friend over. She stepped up, and it worked instantly. I tried again, standing closer, adjusting my position. Still nothing. Then another friend with a complexion similar to mine approached. The machine failed for her, too.</p><p>That’s when it hit me: the facial recognition system didn’t work consistently on darker skin tones. In our group, the pattern was unmistakable. It only failed for those of us with darker skin.</p><p>My initial reaction was disbelief. How could cutting-edge technology designed for public safety exclude entire groups of people? But here’s what I’ve learned: this wasn’t intentional discrimination. It was something more subtle and, in many ways, more concerning. The AI simply didn’t know how to see us.</p><h3>The Invisible Problem in the Data</h3><figure><img alt="Infographic that outlines how to identify and stop AI Bias. The image depicts the flow of where the bias begins and how it is integrated into the AI models. The bias can begin at data collection then end up in model training and subsequently to deployment. The infographic also outlines ways to prevent or reduce AI Bias, through fairness audits, involving diverse stakeholders and create explainable and accountable AI." src="https://cdn-images-1.medium.com/max/1024/1*sCvxpC0pPnrhcp7x0nxIsA.png" /><figcaption>Source: <a href="https://onix-systems.com/blog/ai-bias-detection-and-mitigation">https://onix-systems.com/blog/ai-bias-detection-and-mitigation</a> Infographic: Notebook LM</figcaption></figure><p>Here’s how it works: AI systems learn from examples, like a child learning to recognize faces by looking at family photos. But what if all those photos showed only one type of face? The child would struggle to recognize faces that looked different.</p><p>That’s exactly what happens with AI. Studies from MIT and other institutions have shown that facial recognition error rates for dark-skinned faces can be dozens of times higher than for lighter-skinned faces. Why? Because the training data (the “photos” the AI learned from) was skewed heavily toward light-skinned faces. The AI literally knows those features better because it’s seen them more.</p><p>In 2022, USC researchers found bias in up to 38.6% of facts used by AI systems. These aren’t just technical glitches. There are knowledge gaps that lead to real consequences for real people.</p><h3>When Bias Affects Your Life</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*F-4gtUjr1mxthnbyWvvHpA.png" /><figcaption>Source: <a href="https://news.stanford.edu/stories/2025/10/ai-llms-age-bias-older-working-women-research">https://news.stanford.edu/stories/2025/10/ai-llms-age-bias-older-working-women-research</a> Infographic: NotebookLM</figcaption></figure><p>The facial recognition exhibit was frustrating, but imagine if that same technology were deciding whether you got a job interview, a loan, or medical treatment.</p><p><strong>In lending:</strong> You have good credit and a steady income, but you live in a certain ZIP code. An AI system flags your application as “high risk” not because of your financial behavior, but because the algorithm learned to associate your neighborhood with defaults. You’re denied a loan you qualified for. A 2019 UC Berkeley study found that lending algorithms were charging Latino and African American borrowers $765 million more annually in interest. That biased data from 2019 and all the years before it is now in the databases that train today’s AI systems, teaching them that this discrimination is normal.</p><p><strong>In hiring:</strong> Your resume gets filtered out before a human ever sees it. Maybe your name sounds “foreign,” or you went to a state school instead of an Ivy League university, or your work history has gaps the AI can’t contextualize. A Stanford study from October 2025 found that AI hiring tools consistently rated older women lower than older men with identical qualifications.</p><p><strong>In healthcare, a widely </strong>used algorithm systematically recommended less care for Black patients than white patients with the same health conditions. The AI wasn’t programmed to discriminate. It was using healthcare costs as a proxy for health needs. But because Black patients historically had less access to healthcare and therefore lower costs, the algorithm incorrectly concluded they were healthier.</p><p>These aren’t isolated incidents. They’re patterns that emerge when AI learns from a world that already has inequalities baked into it.</p><h3>The 2026 Reality Check</h3><p>You might think, “Surely we’ve fixed this by now?” Unfortunately, recent research shows the problem is evolving, not disappearing.</p><p>A July 2025 study published in the Proceedings of the National Academy of Sciences discovered something troubling: AI systems now prefer content created by other AI over content created by humans, even when human evaluators can’t tell the difference. As AI increasingly helps make hiring and purchasing decisions, people who don’t use AI assistance may be systematically disadvantaged.</p><p>New regulations are taking effect in 2026. California, Colorado, and the EU all have laws requiring companies to test for bias and document their AI systems. But here’s the gap: while 64% of organizations say they’re concerned about AI bias, only 58% have even assessed their AI risks. And 42% of companies abandoned most of their AI initiatives in 2025 because they couldn’t get them to work reliably.</p><h3>What We Can Do</h3><p>The good news? This isn’t unfixable. Because AI bias often comes from incomplete or skewed training data rather than malicious intent, we can address it by:</p><p><strong>Diversifying the data:</strong> Making sure AI systems learn from examples that represent all the people they’ll serve</p><p><strong>Testing across groups:</strong> Checking if the AI performs equally well for different demographics before deployment</p><p><strong>Maintaining human oversight:</strong> Keeping humans in the loop for high-stakes decisions.</p><p><strong>Demanding transparency:</strong> Requiring companies to show us how their AI systems work and who they might disadvantage.</p><p>Standing in that exhibit, unable to be recognized by a system meant to represent our future, I realized something important: technology isn’t neutral just because it’s math and code. It reflects the choices of what data to include, what to measure, and who gets to decide.</p><p>The future I encountered on that exhibit floor didn’t see me. But we have the power to build AI systems that do see everyone if we demand it, test for it, and refuse to accept anything less.</p><p>Because the cost of getting this wrong isn’t just a broken exhibit. It’s opportunities denied, care withheld, and inequality amplified under the guise of objectivity.</p><p>What’s your experience with AI bias? Share your story in the comments below.”</p><p><strong><em>Disclosure: </em></strong><em>“This article was written with AI assistance to help research and edit grammar of the article. All experiences, perspectives, and conclusions are my own.”</em></p><p><strong>Key Sources:</strong></p><p>USC Viterbi School of Engineering (2022): Found bias in up to 38.6% of facts used by AI</p><p>Buolamwini &amp; Gebru (2018): MIT study showing facial recognition error rates dozens of times higher for darker-skinned faces</p><p>UC Berkeley (2019): Study found lending algorithms charged Latino and African American borrowers $765 million more annually in interest</p><p>Stanford University (October 2025): AI bias against older women in hiring</p><p>PNAS (July 2025): AI systems preferring AI-generated content over human content</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=3ffdc37b6dc0" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[When a recruiter told me only 33% of My Career Mattered]]></title>
            <description><![CDATA[<div class="medium-feed-item"><p class="medium-feed-image"><a href="https://medium.com/@delethia/when-a-recruiter-told-me-only-33-of-my-career-mattered-63611c3db743?source=rss-cb64445e7d1a------2"><img src="https://cdn-images-1.medium.com/max/1536/1*TlG_QG0b2JF13fFiRI4FBw.png" width="1536"></a></p><p class="medium-feed-snippet">Subtitle: After 30 years of building systems, leading teams, and surviving recessions, I was told to erase two decades of experience&#x2026;</p><p class="medium-feed-link"><a href="https://medium.com/@delethia/when-a-recruiter-told-me-only-33-of-my-career-mattered-63611c3db743?source=rss-cb64445e7d1a------2">Continue reading on Medium »</a></p></div>]]></description>
            <link>https://medium.com/@delethia/when-a-recruiter-told-me-only-33-of-my-career-mattered-63611c3db743?source=rss-cb64445e7d1a------2</link>
            <guid isPermaLink="false">https://medium.com/p/63611c3db743</guid>
            <category><![CDATA[career-change]]></category>
            <category><![CDATA[ageism]]></category>
            <category><![CDATA[génération-x]]></category>
            <category><![CDATA[layoffs]]></category>
            <category><![CDATA[personal-growth]]></category>
            <dc:creator><![CDATA[Delethia Johnson]]></dc:creator>
            <pubDate>Tue, 30 Sep 2025 02:59:08 GMT</pubDate>
            <atom:updated>2025-09-30T02:59:08.674Z</atom:updated>
        </item>
        <item>
            <title><![CDATA[Rest, Reflect, Reset]]></title>
            <link>https://medium.com/@delethia/rest-reflect-reset-fe0d3414d75a?source=rss-cb64445e7d1a------2</link>
            <guid isPermaLink="false">https://medium.com/p/fe0d3414d75a</guid>
            <category><![CDATA[rest]]></category>
            <category><![CDATA[new-beginnings]]></category>
            <category><![CDATA[entrepreneurship]]></category>
            <dc:creator><![CDATA[Delethia Johnson]]></dc:creator>
            <pubDate>Sun, 17 Sep 2023 21:23:41 GMT</pubDate>
            <atom:updated>2023-09-17T21:23:41.156Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*orzxN3NH8f6nX7wo" /><figcaption>Photo by <a href="https://unsplash.com/@sincerelymedia?utm_source=medium&amp;utm_medium=referral">Sincerely Media</a> on <a href="https://unsplash.com/?utm_source=medium&amp;utm_medium=referral">Unsplash</a></figcaption></figure><p>This blog might look a little different from what you usually read on Medium about entrepreneurship. I don’t have an earnings story to capture your interest or some new breakthrough to reveal on making money online. Well, at least not yet. You see, I am new here on Medium, and this is my first article.</p><p>A good friend of mine encouraged me to write. The last year of my life has been rough, and I have been wearing a mask all this time. She thought writing would help me, and someone reading this story could relate and benefit from it, so here I am. I decided to write on Medium as I hope this story will inspire you to pursue your dreams and work towards living your best life.</p><p>So here’s my story. I went through a rough breakup a year ago, and the timing of this breakup was great. Ha, I am being facetious. In August of 2022, I started a transportation business with a man I thought would be my husband. Six weeks later, on a Saturday in September, things changed. I received a phone call that disrupted everything for me. I discovered that I was participating in a group activity that was unknown to me. Yes, he had been seeing another woman when we were dating. Once I received that phone call, so many other things came out, and like that, the relationship was over, but we had this business. Now, what am I going to do? Any relationship has ups and downs, and everything is much clearer when looking back. You know that old saying, “Hindsight is twenty-twenty.”</p><p>Needless to say, I was devastated. I felt conned. Fast forward, the business is dissolved, but I took a hit financially. I took a hit mentally. I couldn’t let this stop me. Although I was broken mentally and financially, I had to keep moving. You see, during that time, at my day job, I was also voluntold to work on a special project that kept me on the road for months. Back-to-back traveling that took a toll on me.</p><p>Every day, I had to stand in front of a room of 30 people with another 30 or so people on the phone. All eyes were on me. I couldn’t break down, I couldn’t lose focus. I had to perform. I had to be at the top of my game. After all, my job counted on it. This project was beyond awful… Is there something worse than awful? I didn’t have a choice, or at least that was how I felt.</p><p>That project finished on July 14th, and that was also my last day in the office before going on leave. I had hip replacement surgery on July 17th. I needed this surgery years ago, but it was canceled in 2020 due to Covid, and then I changed my mind about surgery, I decided I wasn’t ready. However, in 2023, the time had come when I couldn’t stand the pain anymore, and I knew I needed a break. Before surgery, people kept telling me I must be happy because I would no longer be in pain. Uh wrong!! I was ecstatic about surgery because I was going on leave and would be away from work. Not being in pain was just a byproduct of what I needed most. I needed to get away from that toxic environment. For months, I experienced workplace bullying and was tired of fighting.</p><p>This leave has been the best experience of my life. I know it might sound crazy, but it has been great. Yes, I was in pain for almost three weeks, and yes, parts of my leg are still numb, but this is the first time in a year that I have had the opportunity to rest, reflect and reset. I took my mask off. I no longer had to smile when I felt like crying. I no longer had to encourage my team when I needed to be encouraged myself. I no longer had to pour into others when I needed to be poured into. I finally had the opportunity to mourn that relationship. Now, I see clearer than ever before about my life and what I want.</p><p>During this time of leave and rest, I realized I was tired of working in corporate. Twenty-eight years in this environment is enough. I am tired of people telling me how much I am worth. I am tired of losing out on jobs to their friends, whether they are qualified or not.</p><p>This leave has motivated me to plan my exit from the corporate space. I am finally finding my voice at 50. I am going through a shift, and I am all for it. I have learned that it’s never too late to start over. It’s never too late to rebuild. It’s too late when you’re gone.</p><p>Wearing a mask can wear you out, and I hope it doesn’t take surgery for you to pursue your dreams. Although I must return to work at some point, I am returning with the mindset that this job is temporary. I am using this leave to develop my exit strategy and plant the seeds for my entrepreneurial success. I don’t know how this will turn out, but you never know until you try. If there is something you are thinking about starting, walk into it and trust the process.</p><p>I encourage anyone reading this to take the time you need to rest, reflect, and reset. I hope you will join me on this new journey. I am excited and nervous at the same time, but what I love about the community on Medium is that there are so many people writing about their successes that I am encouraged that I will have a success story to share with you soon.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=fe0d3414d75a" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>