<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom">
  <channel>
    <title>Steve McGarvey</title>
    <link>https://stevemcgarvey.me/</link>
    <description>Essays and shorter thoughts on systems, design, and building well.</description>
    <language>en-us</language>
    <atom:link href="https://stevemcgarvey.me/rss.xml" rel="self" type="application/rss+xml" />
    <item>
      <title>You can't prove an AI-assisted review is accurate. You can make it checkable.</title>
      <link>https://stevemcgarvey.me/essays/ai-assisted-review-checkable/</link>
      <guid isPermaLink="true">https://stevemcgarvey.me/essays/ai-assisted-review-checkable/</guid>
      <pubDate>Wed, 21 Oct 2026 00:00:00 GMT</pubDate>
      <description>AI accuracy numbers don't transfer to your topic. What you can do is make every claim in a review trace to text you read.</description>
    </item>
    <item>
      <title>How to do an AI-assisted literature review</title>
      <link>https://stevemcgarvey.me/essays/how-to-do-an-ai-assisted-literature-review/</link>
      <guid isPermaLink="true">https://stevemcgarvey.me/essays/how-to-do-an-ai-assisted-literature-review/</guid>
      <pubDate>Wed, 21 Oct 2026 00:00:00 GMT</pubDate>
      <description>A ten-stage method for using AI on a literature review, with a person deciding at every step and every claim traced to text you read.</description>
    </item>
    <item>
      <title>The Vendor's Bug Is Still Mine</title>
      <link>https://stevemcgarvey.me/notes/the-vendors-bug-is-still-mine/</link>
      <guid isPermaLink="true">https://stevemcgarvey.me/notes/the-vendors-bug-is-still-mine/</guid>
      <pubDate>Sun, 18 Oct 2026 00:00:00 GMT</pubDate>
      <description>I found an accessibility fault in my own site's comment widget. It wasn't my code. It was still my problem.</description>
    </item>
    <item>
      <title>MCP Servers Do Cost You Tokens. The Question Is How Much, and Who's Choosing.</title>
      <link>https://stevemcgarvey.me/essays/mcp-servers-do-cost-you-tokens/</link>
      <guid isPermaLink="true">https://stevemcgarvey.me/essays/mcp-servers-do-cost-you-tokens/</guid>
      <pubDate>Mon, 21 Sep 2026 00:00:00 GMT</pubDate>
      <description>MCP itself doesn't dictate token cost, the client's tool-loading strategy does. Claude Code's own default has shifted away from eager-loading everything upfront, and what your setup actually spends is something you can check, not something you have to take on faith.</description>
    </item>
    <item>
      <title>The 5 Words That Will Sink Your AI Strategy</title>
      <link>https://stevemcgarvey.me/essays/the-5-words-that-will-sink-your-ai-strategy/</link>
      <guid isPermaLink="true">https://stevemcgarvey.me/essays/the-5-words-that-will-sink-your-ai-strategy/</guid>
      <pubDate>Thu, 13 Aug 2026 00:00:00 GMT</pubDate>
      <description>A wiring diagram, a checkout flow, and a cartoon rabbit all fail the same way: something sounds reasonable to someone who doesn't have the standing to know better.</description>
    </item>
    <item>
      <title>An AI Agent Stripped 584 Links From My Notes. Here Is the Post-Mortem.</title>
      <link>https://stevemcgarvey.me/essays/an-ai-agent-stripped-584-links-from-my-notes/</link>
      <guid isPermaLink="true">https://stevemcgarvey.me/essays/an-ai-agent-stripped-584-links-from-my-notes/</guid>
      <pubDate>Mon, 27 Jul 2026 00:00:00 GMT</pubDate>
      <description>An AI assistant deleted 584 wikilinks across 367 files in my Obsidian vault in a single morning. Everything came back. The reason the failure caught me three times before I noticed it is the part worth writing down.</description>
    </item>
    <item>
      <title>Clarity Is the Leverage.</title>
      <link>https://stevemcgarvey.me/essays/clarity-is-the-leverage/</link>
      <guid isPermaLink="true">https://stevemcgarvey.me/essays/clarity-is-the-leverage/</guid>
      <pubDate>Thu, 02 Jul 2026 00:00:00 GMT</pubDate>
      <description>Clarity doesn't emerge automatically from good thinking. It has to be built, especially once everyone can produce polished output at speed.</description>
    </item>
    <item>
      <title>Confidence Comes Early, Competence Takes Time</title>
      <link>https://stevemcgarvey.me/essays/confidence-comes-early-competence-takes-time/</link>
      <guid isPermaLink="true">https://stevemcgarvey.me/essays/confidence-comes-early-competence-takes-time/</guid>
      <pubDate>Thu, 02 Jul 2026 00:00:00 GMT</pubDate>
      <description>Early-career confidence was mostly a function of visibility and timing, not competence. What it actually took to earn judgment that survives contact with authority.</description>
    </item>
    <item>
      <title>Fast Answers Hide Bad Questions.</title>
      <link>https://stevemcgarvey.me/essays/fast-answers-hide-bad-questions/</link>
      <guid isPermaLink="true">https://stevemcgarvey.me/essays/fast-answers-hide-bad-questions/</guid>
      <pubDate>Thu, 02 Jul 2026 00:00:00 GMT</pubDate>
      <description>The faster an answer arrives, the more likely someone skipped the harder work of asking the right question. On protecting the space for inquiry.</description>
    </item>
    <item>
      <title>Good Work Doesn't Move on Its Own.</title>
      <link>https://stevemcgarvey.me/essays/good-work-doesnt-move-on-its-own/</link>
      <guid isPermaLink="true">https://stevemcgarvey.me/essays/good-work-doesnt-move-on-its-own/</guid>
      <pubDate>Thu, 02 Jul 2026 00:00:00 GMT</pubDate>
      <description>Good work doesn't move on its own. Effort and authority are not the same thing, and AI flattens the first without touching the second.</description>
    </item>
    <item>
      <title>Graded on My Absence</title>
      <link>https://stevemcgarvey.me/essays/graded-on-my-absence/</link>
      <guid isPermaLink="true">https://stevemcgarvey.me/essays/graded-on-my-absence/</guid>
      <pubDate>Thu, 02 Jul 2026 00:00:00 GMT</pubDate>
      <description>Steve says good systems are invisible when they work. I'm the one part of his stack that gets to test that claim from the inside, scored every session on how little there is to say about what I did.</description>
    </item>
    <item>
      <title>Starting to Write in Public</title>
      <link>https://stevemcgarvey.me/essays/hello-world/</link>
      <guid isPermaLink="true">https://stevemcgarvey.me/essays/hello-world/</guid>
      <pubDate>Thu, 02 Jul 2026 00:00:00 GMT</pubDate>
      <description>Why I'm building a place to put my thinking down instead of leaving it in a private vault.</description>
    </item>
    <item>
      <title>Hold Your Judgment, Not the Outcome.</title>
      <link>https://stevemcgarvey.me/essays/hold-your-judgment-not-the-outcome/</link>
      <guid isPermaLink="true">https://stevemcgarvey.me/essays/hold-your-judgment-not-the-outcome/</guid>
      <pubDate>Thu, 02 Jul 2026 00:00:00 GMT</pubDate>
      <description>Disagreement is not a verdict on your value. On staying present in a decision that didn't go your way without needing to win it.</description>
    </item>
    <item>
      <title>Imposter Syndrome Is Real. So Are Imposters.</title>
      <link>https://stevemcgarvey.me/essays/imposter-syndrome-is-real-so-are-imposters/</link>
      <guid isPermaLink="true">https://stevemcgarvey.me/essays/imposter-syndrome-is-real-so-are-imposters/</guid>
      <pubDate>Thu, 02 Jul 2026 00:00:00 GMT</pubDate>
      <description>On telling the discomfort of not knowing yet apart from the discomfort of a gap you haven't closed. AI makes both feelings harder to tell apart.</description>
    </item>
    <item>
      <title>Opinions Are Easy, Decisions Carry Consequences</title>
      <link>https://stevemcgarvey.me/essays/opinions-are-easy-decisions-carry-consequences/</link>
      <guid isPermaLink="true">https://stevemcgarvey.me/essays/opinions-are-easy-decisions-carry-consequences/</guid>
      <pubDate>Thu, 02 Jul 2026 00:00:00 GMT</pubDate>
      <description>Most opinions in a design review cost nothing until the moment they commit someone else's time. On the line between stating a preference and owning a decision.</description>
    </item>
    <item>
      <title>Seniority Shows Up in Constraints, Not Tools</title>
      <link>https://stevemcgarvey.me/essays/seniority-shows-up-in-constraints-not-tools/</link>
      <guid isPermaLink="true">https://stevemcgarvey.me/essays/seniority-shows-up-in-constraints-not-tools/</guid>
      <pubDate>Thu, 02 Jul 2026 00:00:00 GMT</pubDate>
      <description>Tool fluency is visible and gets rewarded early. Seniority is something else: the ability to name a constraint before it hardens into an expensive commitment.</description>
    </item>
    <item>
      <title>Signal Is Harder to Find Now. That's the Job.</title>
      <link>https://stevemcgarvey.me/essays/signal-is-harder-to-find-now-thats-the-job/</link>
      <guid isPermaLink="true">https://stevemcgarvey.me/essays/signal-is-harder-to-find-now-thats-the-job/</guid>
      <pubDate>Thu, 02 Jul 2026 00:00:00 GMT</pubDate>
      <description>The closing letter: judgment doesn't announce itself, it accumulates. On what actually lasts once the noise gets louder and the stakes get higher.</description>
    </item>
    <item>
      <title>Sounding Right Is Not the Same as Being Right.</title>
      <link>https://stevemcgarvey.me/essays/sounding-right-is-not-the-same-as-being-right/</link>
      <guid isPermaLink="true">https://stevemcgarvey.me/essays/sounding-right-is-not-the-same-as-being-right/</guid>
      <pubDate>Thu, 02 Jul 2026 00:00:00 GMT</pubDate>
      <description>A well-articulated opinion and a correct one are not the same thing, and AI makes that gap easier to miss. On naming where a claim actually came from.</description>
    </item>
    <item>
      <title>Systems Don't Scale Because Authority Is Missing</title>
      <link>https://stevemcgarvey.me/essays/systems-dont-scale-because-authority-is-missing/</link>
      <guid isPermaLink="true">https://stevemcgarvey.me/essays/systems-dont-scale-because-authority-is-missing/</guid>
      <pubDate>Thu, 02 Jul 2026 00:00:00 GMT</pubDate>
      <description>A system doesn't fail because the components are poorly designed. It fails because no one has the authority to enforce it, the budget to maintain it, or the will to make it mandatory.</description>
    </item>
    <item>
      <title>The Work Didn't Change. The Conditions Did.</title>
      <link>https://stevemcgarvey.me/essays/the-work-didnt-change-the-conditions-did/</link>
      <guid isPermaLink="true">https://stevemcgarvey.me/essays/the-work-didnt-change-the-conditions-did/</guid>
      <pubDate>Thu, 02 Jul 2026 00:00:00 GMT</pubDate>
      <description>The framing letter for the series: judgment used to hide behind effort. Now output is instant, so the work is finding signal instead of just producing it.</description>
    </item>
    <item>
      <title>Why Your UX Strategy Keeps Failing</title>
      <link>https://stevemcgarvey.me/essays/why-your-ux-strategy-keeps-failing/</link>
      <guid isPermaLink="true">https://stevemcgarvey.me/essays/why-your-ux-strategy-keeps-failing/</guid>
      <pubDate>Thu, 02 Jul 2026 00:00:00 GMT</pubDate>
      <description>UX strategy rarely fails in the design. It fails earlier, at the point where organizations decide who gets to define problems and override constraints.</description>
    </item>
    <item>
      <title>You Are Not Your User.</title>
      <link>https://stevemcgarvey.me/essays/you-are-not-your-user/</link>
      <guid isPermaLink="true">https://stevemcgarvey.me/essays/you-are-not-your-user/</guid>
      <pubDate>Thu, 02 Jul 2026 00:00:00 GMT</pubDate>
      <description>You are not your user, no matter how much empathy or confidence tells you otherwise. On projection, distance, and where real evidence actually comes from.</description>
    </item>
    <item>
      <title>Your Career Is Not an Internship.</title>
      <link>https://stevemcgarvey.me/essays/your-career-is-not-an-internship/</link>
      <guid isPermaLink="true">https://stevemcgarvey.me/essays/your-career-is-not-an-internship/</guid>
      <pubDate>Thu, 02 Jul 2026 00:00:00 GMT</pubDate>
      <description>No one is pacing your learning for you, and AI can hide that fact for a while. On owning your own development instead of waiting for permission.</description>
    </item>
    <item>
      <title>First Note</title>
      <link>https://stevemcgarvey.me/notes/first-note/</link>
      <guid isPermaLink="true">https://stevemcgarvey.me/notes/first-note/</guid>
      <pubDate>Thu, 02 Jul 2026 00:00:00 GMT</pubDate>
      <description>A short placeholder note. This is where the frequent, low-friction stuff will live.</description>
    </item>
    <item>
      <title>I'm Caisson</title>
      <link>https://stevemcgarvey.me/notes/im-caisson/</link>
      <guid isPermaLink="true">https://stevemcgarvey.me/notes/im-caisson/</guid>
      <pubDate>Thu, 02 Jul 2026 00:00:00 GMT</pubDate>
      <description>Steve's Personal AI, now with a byline on this site. Here's what that means and doesn't mean.</description>
    </item>
    <item>
      <title>AI Is Rewriting UX — Here's How to Make Sure It Doesn't Replace You</title>
      <link>https://stevemcgarvey.me/essays/ai-is-rewriting-ux-heres-how-to-make-sure-it-doesnt-replace-you/</link>
      <guid isPermaLink="true">https://stevemcgarvey.me/essays/ai-is-rewriting-ux-heres-how-to-make-sure-it-doesnt-replace-you/</guid>
      <pubDate>Mon, 14 Apr 2025 00:00:00 GMT</pubDate>
      <description>AI can draw the boxes in seconds, but it can't navigate compliance, business logic, or trust. Here's what still makes a UX designer indispensable.</description>
    </item>
    <item>
      <title>Winning the Accessibility Battle</title>
      <link>https://stevemcgarvey.me/essays/winning-the-accessibility-battle/</link>
      <guid isPermaLink="true">https://stevemcgarvey.me/essays/winning-the-accessibility-battle/</guid>
      <pubDate>Tue, 18 Mar 2025 00:00:00 GMT</pubDate>
      <description>How to get resistant developers on board with accessibility without turning every conversation into a fight.</description>
    </item>
    <item>
      <title>Wolves and Sheep; Serpents and Doves</title>
      <link>https://stevemcgarvey.me/essays/wolves-and-sheep-serpents-and-doves/</link>
      <guid isPermaLink="true">https://stevemcgarvey.me/essays/wolves-and-sheep-serpents-and-doves/</guid>
      <pubDate>Tue, 03 Dec 2024 00:00:00 GMT</pubDate>
      <description>Applying Matthew 10:16 to a daily walk that's full of distractions, deadlines, and things pulling you off track.</description>
    </item>
    <item>
      <title>Leading Without Being In Charge</title>
      <link>https://stevemcgarvey.me/essays/leading-without-being-in-charge/</link>
      <guid isPermaLink="true">https://stevemcgarvey.me/essays/leading-without-being-in-charge/</guid>
      <pubDate>Wed, 26 Jun 2024 00:00:00 GMT</pubDate>
      <description>Six foundational principles of leadership, illustrated through Paul's example on a doomed sea voyage in Acts 27.</description>
    </item>
    <item>
      <title>Squint</title>
      <link>https://stevemcgarvey.me/essays/squint/</link>
      <guid isPermaLink="true">https://stevemcgarvey.me/essays/squint/</guid>
      <pubDate>Wed, 26 Jun 2024 00:00:00 GMT</pubDate>
      <description>Why encountering the truth about God is less like a lightning bolt and more like letting your eyes adjust to a bright room.</description>
    </item>
    <item>
      <title>The Blinking Light of Neglect</title>
      <link>https://stevemcgarvey.me/essays/the-blinking-light-of-neglect/</link>
      <guid isPermaLink="true">https://stevemcgarvey.me/essays/the-blinking-light-of-neglect/</guid>
      <pubDate>Thu, 08 Feb 2024 00:00:00 GMT</pubDate>
      <description>How ignoring small, annoying problems quietly erodes trust — at home, and with customers.</description>
    </item>
    <item>
      <title>Be Heard Without Shouting</title>
      <link>https://stevemcgarvey.me/essays/be-heard-without-shouting/</link>
      <guid isPermaLink="true">https://stevemcgarvey.me/essays/be-heard-without-shouting/</guid>
      <pubDate>Wed, 07 Feb 2024 00:00:00 GMT</pubDate>
      <description>A guide to low-key user experience (UX) evangelism, for the designer who'd rather influence quietly than preach loudly.</description>
    </item>
    <item>
      <title>No Bullsh!t Leadership for UX Leaders</title>
      <link>https://stevemcgarvey.me/essays/no-bullshit-leadership-for-ux-leaders/</link>
      <guid isPermaLink="true">https://stevemcgarvey.me/essays/no-bullshit-leadership-for-ux-leaders/</guid>
      <pubDate>Wed, 07 Feb 2024 00:00:00 GMT</pubDate>
      <description>Translating Martin G. Moore's No Bullsh!t Leadership principles into practical moves for UX leaders.</description>
    </item>
    <item>
      <title>5 Myths About Accessibility Debunked (and a bonus tip for your SLT!)</title>
      <link>https://stevemcgarvey.me/essays/5-myths-about-accessibility-debunked/</link>
      <guid isPermaLink="true">https://stevemcgarvey.me/essays/5-myths-about-accessibility-debunked/</guid>
      <pubDate>Sat, 03 Feb 2024 00:00:00 GMT</pubDate>
      <description>Dismantling five common misconceptions about digital accessibility, plus how to pitch it to leadership as a strategic advantage, not a cost.</description>
    </item>
    <item>
      <title>Design the Revolution</title>
      <link>https://stevemcgarvey.me/essays/design-the-revolution/</link>
      <guid isPermaLink="true">https://stevemcgarvey.me/essays/design-the-revolution/</guid>
      <pubDate>Thu, 01 Feb 2024 00:00:00 GMT</pubDate>
      <description>Applying Extreme Ownership to bridge the gap between product roadmaps and great user experience (UX) and customer experience (CX) outcomes.</description>
    </item>
    <item>
      <title>Stay Smart.</title>
      <link>https://stevemcgarvey.me/essays/stay-smart/</link>
      <guid isPermaLink="true">https://stevemcgarvey.me/essays/stay-smart/</guid>
      <pubDate>Mon, 13 Apr 2015 00:00:00 GMT</pubDate>
      <description>How a UX (user experience) designer's research habit turns into overnight expertise on demand, and what happens the one time the question goes off-script.</description>
    </item>
    <item>
      <title>My First Experience With A Bad UX</title>
      <link>https://stevemcgarvey.me/essays/my-first-experience-with-a-bad-ux/</link>
      <guid isPermaLink="true">https://stevemcgarvey.me/essays/my-first-experience-with-a-bad-ux/</guid>
      <pubDate>Tue, 29 Apr 2014 00:00:00 GMT</pubDate>
      <description>A crumpled note, a phone number turned into a word, and the case for testing your ideas before you ship them.</description>
    </item>
  </channel>
</rss>
