<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom">
  <channel>
    <title>Matthew Jefferies</title>
    <link>https://matthewjefferies.com</link>
    <description>Where AI acceleration meets institutional inertia. Strategic analysis of how AI reshapes workforce strategy, career trajectories, and institutional capability.</description>
    <language>en-au</language>
    <managingEditor>contact@matthewjefferies.com (Matthew Jefferies)</managingEditor>
    <webMaster>contact@matthewjefferies.com (Matthew Jefferies)</webMaster>
    <lastBuildDate>Sat, 30 May 2026 22:30:39 GMT</lastBuildDate>
    <atom:link href="https://matthewjefferies.com/feed.xml" rel="self" type="application/rss+xml"/>
    <item>
      <title>AI found a way past the front door</title>
      <link>https://matthewjefferies.com/articles/ai-found-a-way-past-the-front-door</link>
      <guid isPermaLink="true">https://matthewjefferies.com/articles/ai-found-a-way-past-the-front-door</guid>
      <description>72% of American teenagers have used AI companions. A third call them friends. Five US states are scrambling to legislate. Australia&apos;s eSafety Commissioner found the platforms are failing. And the adults scrolling past AI slop on LinkedIn have not yet noticed the other half of the problem.</description>
      <pubDate>Tue, 21 Apr 2026 00:00:00 GMT</pubDate>
      <author>contact@matthewjefferies.com (Matthew Jefferies)</author>
      <category>ai acceleration</category>
      <category>institutional adaptation</category>
      <category>workforce strategy</category>
      <category>ai companions</category>
      <category>regulation</category>
      <category>outpaced investigation</category>
    </item>
    <item>
      <title>The Snap signal: AI just became a capital-markets story</title>
      <link>https://matthewjefferies.com/articles/the-snap-signal</link>
      <guid isPermaLink="true">https://matthewjefferies.com/articles/the-snap-signal</guid>
      <description>Snap cut 16% of its workforce and the stock jumped 7%. That is the signal every CFO was waiting for. What happens when the market starts rewarding AI-driven headcount cuts.</description>
      <pubDate>Sat, 18 Apr 2026 00:00:00 GMT</pubDate>
      <author>contact@matthewjefferies.com (Matthew Jefferies)</author>
      <category>workforce strategy</category>
      <category>defence security</category>
      <category>ai layoffs</category>
      <category>workforce disruption</category>
      <category>capital markets</category>
      <category>workforce design</category>
      <category>outpaced investigation</category>
    </item>
    <item>
      <title>Five charts from the Stanford AI Index 2026 that should worry every workforce planner</title>
      <link>https://matthewjefferies.com/articles/stanford-ai-index-2026-workforce-charts</link>
      <guid isPermaLink="true">https://matthewjefferies.com/articles/stanford-ai-index-2026-workforce-charts</guid>
      <description>The 2026 AI Index is the first edition where workforce disruption stops being a forecast and starts being a measurement. Five charts a planner cannot ignore.</description>
      <pubDate>Fri, 17 Apr 2026 00:00:00 GMT</pubDate>
      <author>contact@matthewjefferies.com (Matthew Jefferies)</author>
      <category>ai acceleration</category>
      <category>workforce strategy</category>
      <category>defence security</category>
      <category>ai adoption</category>
      <category>compute infrastructure</category>
      <category>strategic risk</category>
      <category>outpaced investigation</category>
    </item>
    <item>
      <title>The AI leverage gap: why 80% of companies spend on AI and get almost nothing back</title>
      <link>https://matthewjefferies.com/articles/ai-leverage-gap</link>
      <guid isPermaLink="true">https://matthewjefferies.com/articles/ai-leverage-gap</guid>
      <description>PwC&apos;s new study finds 20% of companies capture 74% of AI&apos;s economic value. The same dynamic plays out at the individual level. Here&apos;s what separates the winners.</description>
      <pubDate>Thu, 16 Apr 2026 00:00:00 GMT</pubDate>
      <author>contact@matthewjefferies.com (Matthew Jefferies)</author>
      <category>ai acceleration</category>
      <category>mental models</category>
      <category>implementation</category>
    </item>
    <item>
      <title>The $425 billion assumption</title>
      <link>https://matthewjefferies.com/articles/the-425-billion-assumption</link>
      <guid isPermaLink="true">https://matthewjefferies.com/articles/the-425-billion-assumption</guid>
      <description>Australia&apos;s 2026 National Defence Strategy commits record spending to counter the most dangerous era since World War II. It never asks whether the economy can sustain it.</description>
      <pubDate>Thu, 16 Apr 2026 00:00:00 GMT</pubDate>
      <author>contact@matthewjefferies.com (Matthew Jefferies)</author>
      <category>defence security</category>
      <category>ai acceleration</category>
      <category>workforce strategy</category>
      <category>institutional adaptation</category>
      <category>national security</category>
      <category>defence spending</category>
      <category>fiscal risk</category>
      <category>australia</category>
      <category>outpaced investigation</category>
    </item>
    <item>
      <title>Agents just beat the human baseline at desktop work</title>
      <link>https://matthewjefferies.com/articles/agents-beat-desktop-baseline</link>
      <guid isPermaLink="true">https://matthewjefferies.com/articles/agents-beat-desktop-baseline</guid>
      <description>GPT-5.4 scored 75.0% on OSWorld-Verified. The average human scored 72.4%. A general-purpose AI model now beats the average person at using a computer, and the playbook for what to hand over first is shorter than you think.</description>
      <pubDate>Tue, 14 Apr 2026 00:00:00 GMT</pubDate>
      <author>contact@matthewjefferies.com (Matthew Jefferies)</author>
      <category>ai acceleration</category>
      <category>implementation</category>
    </item>
    <item>
      <title>The AI layoff excuse: 80,000 jobs and the biggest corporate fiction of 2026</title>
      <link>https://matthewjefferies.com/articles/the-ai-layoff-excuse</link>
      <guid isPermaLink="true">https://matthewjefferies.com/articles/the-ai-layoff-excuse</guid>
      <description>Half of Q1&apos;s tech layoffs blame AI. The Federal Reserve, Oxford Economics, and even Sam Altman say that&apos;s mostly nonsense. Here&apos;s what&apos;s actually happening.</description>
      <pubDate>Tue, 14 Apr 2026 00:00:00 GMT</pubDate>
      <author>contact@matthewjefferies.com (Matthew Jefferies)</author>
      <category>workforce strategy</category>
      <category>ai acceleration</category>
      <category>ai layoffs</category>
      <category>workforce disruption</category>
      <category>ai washing</category>
      <category>corporate strategy</category>
      <category>outpaced investigation</category>
    </item>
    <item>
      <title>The three-tool trap</title>
      <link>https://matthewjefferies.com/articles/three-tool-trap</link>
      <guid isPermaLink="true">https://matthewjefferies.com/articles/three-tool-trap</guid>
      <description>The average organisation now runs seven AI tools. The productivity threshold is three. This is a contrarian guide to why more AI tools means less AI productivity, and what to do about it.</description>
      <pubDate>Tue, 14 Apr 2026 00:00:00 GMT</pubDate>
      <author>contact@matthewjefferies.com (Matthew Jefferies)</author>
      <category>ai acceleration</category>
      <category>mental models</category>
      <category>implementation</category>
    </item>
    <item>
      <title>Australia is not ready</title>
      <link>https://matthewjefferies.com/articles/australia-ai-adoption-gap</link>
      <guid isPermaLink="true">https://matthewjefferies.com/articles/australia-ai-adoption-gap</guid>
      <description>Australia released its National AI Plan in December 2025. Singapore&apos;s AI adoption rate is 60.9%. Ours is somewhere around 37% for SMEs. The gap is not closing, and the reasons are structural.</description>
      <pubDate>Sun, 12 Apr 2026 00:00:00 GMT</pubDate>
      <author>contact@matthewjefferies.com (Matthew Jefferies)</author>
      <category>institutional adaptation</category>
      <category>workforce strategy</category>
      <category>defence security</category>
      <category>australia</category>
    </item>
    <item>
      <title>The context window is your competitive advantage</title>
      <link>https://matthewjefferies.com/articles/context-window-competitive-advantage</link>
      <guid isPermaLink="true">https://matthewjefferies.com/articles/context-window-competitive-advantage</guid>
      <description>Most people paste a paragraph into ChatGPT and call it AI. The practitioners rewriting the rules are feeding entire codebases, full research libraries, and complete project histories into a single conversation. The difference is the context window, and almost nobody is using it properly.</description>
      <pubDate>Sun, 12 Apr 2026 00:00:00 GMT</pubDate>
      <author>contact@matthewjefferies.com (Matthew Jefferies)</author>
      <category>mental models</category>
      <category>implementation</category>
    </item>
    <item>
      <title>The six-year collapse</title>
      <link>https://matthewjefferies.com/articles/the-six-year-collapse</link>
      <guid isPermaLink="true">https://matthewjefferies.com/articles/the-six-year-collapse</guid>
      <description>Cognizant&apos;s updated workforce study shows AI disruption originally projected for 2032 is already here. The institutions that planned for a gradual transition just ran out of time.</description>
      <pubDate>Mon, 30 Mar 2026 00:00:00 GMT</pubDate>
      <author>contact@matthewjefferies.com (Matthew Jefferies)</author>
      <category>ai acceleration</category>
      <category>workforce strategy</category>
      <category>defence security</category>
      <category>institutional adaptation</category>
      <category>strategic risk</category>
      <category>institutional lag</category>
    </item>
    <item>
      <title>The pilot phase is over</title>
      <link>https://matthewjefferies.com/articles/agentic-ai-enterprise-production</link>
      <guid isPermaLink="true">https://matthewjefferies.com/articles/agentic-ai-enterprise-production</guid>
      <description>61% of global CEOs are actively deploying AI agents. Three-quarters of enterprises plan agentic AI within two years. The governance infrastructure has not kept pace. Here is what that means.</description>
      <pubDate>Fri, 27 Mar 2026 00:00:00 GMT</pubDate>
      <author>contact@matthewjefferies.com (Matthew Jefferies)</author>
      <category>ai acceleration</category>
      <category>institutional adaptation</category>
      <category>implementation</category>
    </item>
    <item>
      <title>Working a six-day week</title>
      <link>https://matthewjefferies.com/articles/ai-workforce-inflection-2026</link>
      <guid isPermaLink="true">https://matthewjefferies.com/articles/ai-workforce-inflection-2026</guid>
      <description>Workers using AI save 7.5 hours per week. HR leaders know it. The job market knows it. The question is whether you do, and what you plan to do about it.</description>
      <pubDate>Fri, 27 Mar 2026 00:00:00 GMT</pubDate>
      <author>contact@matthewjefferies.com (Matthew Jefferies)</author>
      <category>workforce strategy</category>
      <category>ai acceleration</category>
      <category>implementation</category>
    </item>
    <item>
      <title>One factory to rule them all</title>
      <link>https://matthewjefferies.com/articles/one-factory-to-rule-them-all</link>
      <guid isPermaLink="true">https://matthewjefferies.com/articles/one-factory-to-rule-them-all</guid>
      <description>Elon Musk announced a chip factory targeting 50x current global output. Most people heard &apos;factory&apos; and moved on. They shouldn&apos;t have.</description>
      <pubDate>Thu, 26 Mar 2026 00:00:00 GMT</pubDate>
      <author>contact@matthewjefferies.com (Matthew Jefferies)</author>
      <category>ai acceleration</category>
      <category>defence security</category>
      <category>semiconductors</category>
      <category>compute infrastructure</category>
      <category>terafab</category>
      <category>tesla</category>
      <category>spacex</category>
      <category>xai</category>
      <category>national security</category>
    </item>
    <item>
      <title>Why mental models matter more than prompts</title>
      <link>https://matthewjefferies.com/articles/why-mental-models-matter</link>
      <guid isPermaLink="true">https://matthewjefferies.com/articles/why-mental-models-matter</guid>
      <description>The difference between AI passengers and AI practitioners is not prompt engineering. It is how they think about the technology.</description>
      <pubDate>Fri, 20 Mar 2026 00:00:00 GMT</pubDate>
      <author>contact@matthewjefferies.com (Matthew Jefferies)</author>
      <category>mental models</category>
      <category>implementation</category>
      <category>ai acceleration</category>
    </item>
  </channel>
</rss>