<?xml version="1.0" encoding="utf-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>fragmented sentences</title><link>https://www.jimgumbley.com/blog/</link><description>Jim Gumbley - AI, cyber, software and the zeitgeist.</description><atom:link href="https://www.jimgumbley.com/blog/feeds/all.rss.xml" rel="self"/><lastBuildDate>Sat, 19 Sep 2026 00:00:00 +0100</lastBuildDate><item><title>Did OpenAI's model want to escape? Or did they train it to?</title><link>https://www.jimgumbley.com/blog/did-openais-model-want-to-escape.html</link><description>&lt;p&gt;Jailbreaking, prompt injection and persistence offer a mechanical explanation for an alarming OpenAI disclosure, with plenty still to worry about.&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Jim Gumbley</dc:creator><pubDate>Sat, 19 Sep 2026 00:00:00 +0100</pubDate><guid>tag:www.jimgumbley.com,2026-09-19:/blog/did-openais-model-want-to-escape.html</guid><category>AI</category><category>openai</category><category>ai safety</category><category>cybersecurity</category><category>jailbreaking</category><category>prompt injection</category></item><item><title>If AI destroys our world, why call it “self-improvement”?</title><link>https://www.jimgumbley.com/blog/if-ai-destroys-our-world-why-call-it-self-improvement.html</link><description>&lt;p&gt;Better at what, and for whom? Keeping AI useful and under human control has to be part of what we mean by improvement.&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Jim Gumbley</dc:creator><pubDate>Sat, 12 Sep 2026 00:00:00 +0100</pubDate><guid>tag:www.jimgumbley.com,2026-09-12:/blog/if-ai-destroys-our-world-why-call-it-self-improvement.html</guid><category>AI</category><category>ai</category><category>alignment</category><category>safety</category><category>recursive self-improvement</category></item><item><title>The timing tradeoff: Sam, Dario and Sacks decisions and the Skynet timeline</title><link>https://www.jimgumbley.com/blog/timing-tradeoff-skynet-timeline.html</link><description>&lt;p&gt;Sam, Dario and Sacks are making consequential decisions about pace, oversight and whose risks count.&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Jim Gumbley</dc:creator><pubDate>Sat, 05 Sep 2026 00:00:00 +0100</pubDate><guid>tag:www.jimgumbley.com,2026-09-05:/blog/timing-tradeoff-skynet-timeline.html</guid><category>AI</category><category>artificial intelligence</category><category>AI safety</category><category>AI governance</category><category>OpenAI</category><category>Anthropic</category></item><item><title>The Code's the Thing</title><link>https://www.jimgumbley.com/blog/the-codes-the-thing.html</link><description>&lt;p&gt;Why working software, not documentation or ceremony, is the artifact that actually reduces uncertainty — and why that lesson still holds in the AI era.&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Jim Gumbley</dc:creator><pubDate>Fri, 22 May 2026 00:00:00 +0100</pubDate><guid>tag:www.jimgumbley.com,2026-05-22:/blog/the-codes-the-thing.html</guid><category>Software</category><category>agile</category><category>software</category><category>management</category><category>ai</category><category>delivery</category></item><item><title>My vision for the future of AI: Anthropic Interview</title><link>https://www.jimgumbley.com/blog/anthropic-interview-ai-vision.html</link><description>&lt;p&gt;Reflections from an Anthropic research interview on why language models need honest framing, and why human agency has to stay at the centre.&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Jim Gumbley</dc:creator><pubDate>Sun, 07 Dec 2025 00:00:00 +0000</pubDate><guid>tag:www.jimgumbley.com,2025-12-07:/blog/anthropic-interview-ai-vision.html</guid><category>AI</category><category>anthropic</category><category>ai</category><category>ethics</category><category>language models</category><category>automation</category></item><item><title>Handel's Messiah at Great Malvern Priory</title><link>https://www.jimgumbley.com/blog/handels-messiah-priory.html</link><description>&lt;p&gt;Hearing &lt;em&gt;Messiah&lt;/em&gt; live again sent me down a rabbit hole of counterpoint, class, and why this very Protestant oratorio still works in 2025.&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Jim Gumbley</dc:creator><pubDate>Sat, 06 Dec 2025 00:00:00 +0000</pubDate><guid>tag:www.jimgumbley.com,2025-12-06:/blog/handels-messiah-priory.html</guid><category>Music</category><category>handel</category><category>messiah</category><category>classical music</category><category>baroque</category></item><item><title>Why I Left Spotify in 2025</title><link>https://www.jimgumbley.com/blog/why-i-left-spotify-2025.html</link><description>&lt;p&gt;After fourteen years as a loyal Spotify customer, I finally cancelled my subscription. Here's why the platform's algorithmic direction no longer matches my simple, self-curated listening style.&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Jim Gumbley</dc:creator><pubDate>Sun, 15 Jun 2025 00:00:00 +0100</pubDate><guid>tag:www.jimgumbley.com,2025-06-15:/blog/why-i-left-spotify-2025.html</guid><category>Technology</category><category>spotify</category><category>music</category><category>streaming</category><category>digital ownership</category></item><item><title>It's a Great Time to be a Pen Tester</title><link>https://www.jimgumbley.com/blog/great-time-pen-tester.html</link><description>&lt;p&gt;Exploring how penetration testing has evolved alongside modern software development practices, and why AI-assisted red teaming might be the future of security validation.&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Jim Gumbley</dc:creator><pubDate>Sat, 07 Jun 2025 00:00:00 +0100</pubDate><guid>tag:www.jimgumbley.com,2025-06-07:/blog/great-time-pen-tester.html</guid><category>Security</category><category>penetration testing</category><category>cybersecurity</category></item><item><title>The Coming Wave of AI Agency: A Security Perspective</title><link>https://www.jimgumbley.com/blog/ai-agency-security-perspective.html</link><description>&lt;p&gt;We're not 6 months from AGI; we're 6 months from confronting the security implications of AI systems with real-world agency.&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Jim Gumbley</dc:creator><pubDate>Sun, 23 Mar 2025 00:00:00 +0000</pubDate><guid>tag:www.jimgumbley.com,2025-03-23:/blog/ai-agency-security-perspective.html</guid><category>AI</category><category>artificial intelligence</category><category>security</category><category>risk management</category><category>governance</category></item><item><title>Something changed in the Tech Industry</title><link>https://www.jimgumbley.com/blog/something-changed-tech-industry.html</link><description>&lt;p&gt;The tech industry has shifted from custom builds to piecing together existing components - it's all just plugging Legos now.&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Jim Gumbley</dc:creator><pubDate>Mon, 09 Sep 2024 00:00:00 +0100</pubDate><guid>tag:www.jimgumbley.com,2024-09-09:/blog/something-changed-tech-industry.html</guid><category>Technology</category><category>technology</category><category>cloud computing</category><category>software development</category></item><item><title>Resident Evil in Virtual Reality is actually terrifying</title><link>https://www.jimgumbley.com/blog/resident-evil-vr-terrifying.html</link><description>&lt;p&gt;I thought I was immune to horror games. Then Resident Evil made me take the headset off and have a sit down.&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Jim Gumbley</dc:creator><pubDate>Sun, 05 Mar 2017 00:00:00 +0000</pubDate><guid>tag:www.jimgumbley.com,2017-03-05:/blog/resident-evil-vr-terrifying.html</guid><category>Gaming</category><category>virtual reality</category><category>gaming</category><category>resident evil</category><category>horror</category></item></channel></rss>