<?xml version="1.0" encoding="UTF-8"?><rss version="2.0"><channel><title>AITi — AI commentary with backbone</title><description>AITi publishes one high-quality original AI essay every week: teardowns, industry analysis, people stories and first-hand practice. Opinionated about AI — not a content mill.</description><link>https://www.getaiti.com/</link><language>en-us</language><item><title>One Person, One Agent</title><link>https://www.getaiti.com/en/articles/008-one-agent-per-person/</link><guid isPermaLink="true">https://www.getaiti.com/en/articles/008-one-agent-per-person/</guid><description>From six months of hands-on Hermes to Dots, Muse, and TraeWork: once an agent has long-term memory, files, permissions, tasks, and a runtime, it is no longer just a chat feature.</description><pubDate>Sat, 03 Oct 2026 00:00:00 GMT</pubDate><language>en-us</language><category>AI Agent</category><category>Persistent Agent</category><category>Hermes</category><category>Industry Analysis</category></item><item><title>No Cloud API — I Built a Fully Local Picture-Book Agent with Qwen-Image 2.1</title><link>https://www.getaiti.com/en/articles/007-local-ai-picturebook-agent/</link><guid isPermaLink="true">https://www.getaiti.com/en/articles/007-local-ai-picturebook-agent/</guid><description>I moved the language model, the agent, and image generation entirely onto my own machine: how a local agent painted an 8-page cat picture book and a 7-page Japanese-style short manga all by itself.</description><pubDate>Wed, 23 Sep 2026 00:00:00 GMT</pubDate><language>en-us</language><category>Local AI</category><category>Agent</category><category>Image Generation</category><category>Hands-on Test</category></item><item><title>Qwen3.8-27B Hands-On: 128GB of MacBook Memory Barely Feeds It</title><link>https://www.getaiti.com/en/articles/006-qwen38-27b-mac/</link><guid isPermaLink="true">https://www.getaiti.com/en/articles/006-qwen38-27b-mac/</guid><description>No spec-sheet talk, just lived experience: after an open-source 27B model lived on my MacBook for over a month — 200,000 tokens fed in, two days of real work — these are the honest impressions.</description><pubDate>Thu, 17 Sep 2026 00:00:00 GMT</pubDate><language>en-us</language><category>Local AI</category><category>Open-source models</category><category>Hands-on tests</category></item><item><title>Do We Still Need to Learn to Code?</title><link>https://www.getaiti.com/en/articles/005-programming-in-ai-era/</link><guid isPermaLink="true">https://www.getaiti.com/en/articles/005-programming-in-ai-era/</guid><description>AI writes the code now, but syntax trivia is depreciating while judgment is appreciating — from Karpathy&apos;s vibe coding to my own Second Brain Graph, the syllabus for learning to program changed, it just didn&apos;t get any smaller.</description><pubDate>Fri, 11 Sep 2026 00:00:00 GMT</pubDate><language>en-us</language><category>In-Depth Analysis</category><category>AI-Assisted Programming</category><category>Programming Education</category></item><item><title>AI Remembers You — But Did It Actually Learn?</title><link>https://www.getaiti.com/en/articles/004-memory-or-learning/</link><guid isPermaLink="true">https://www.getaiti.com/en/articles/004-memory-or-learning/</guid><description>AI remembering you is not the same as AI learning you: memory stores the past, real learning has to change the future. Starting from one awkward voice conversation, on memory, reinforcement learning, and the part of a personal AI that actually belongs to the user.</description><pubDate>Mon, 31 Aug 2026 00:00:00 GMT</pubDate><language>en-us</language><category>In-Depth Analysis</category><category>Personal AI</category><category>Reinforcement Learning</category></item><item><title>From &quot;It Runs&quot; to &quot;It Works&quot;: What&apos;s Still Missing in Local AI</title><link>https://www.getaiti.com/en/articles/003-from-cloud-to-edge/</link><guid isPermaLink="true">https://www.getaiti.com/en/articles/003-from-cloud-to-edge/</guid><description>Running isn&apos;t the same as being good to use — models, quantization, inference frameworks, speculative decoding, and agent harnesses are all maturing at once, so local AI&apos;s inflection point isn&apos;t one magic day but an entire stack crossing the usability line.</description><pubDate>Wed, 26 Aug 2026 00:00:00 GMT</pubDate><language>en-us</language><category>In-Depth Analysis</category><category>On-Device AI</category><category>Local AI</category></item><item><title>I Raised a Fully Local AI Companion on My Mac</title><link>https://www.getaiti.com/en/articles/002-local-ai-companion/</link><guid isPermaLink="true">https://www.getaiti.com/en/articles/002-local-ai-companion/</guid><description>One MacBook, one fully local ASR + LLM + TTS + avatar + memory stack: how I took an AI digital companion from demo to something you can actually talk to, and the two product paths it made me rethink.</description><pubDate>Mon, 17 Aug 2026 00:00:00 GMT</pubDate><language>en-us</language><category>local AI</category><category>digital companion</category><category>hands-on test</category></item><item><title>Why Titanium</title><link>https://www.getaiti.com/en/articles/001-why-aiti/</link><guid isPermaLink="true">https://www.getaiti.com/en/articles/001-why-aiti/</guid><description>Why would an AI publication name itself after a metal? AITi&apos;s founding essay explains where the name came from, where the magazine is headed, and what you&apos;ll find here every week.</description><pubDate>Fri, 14 Aug 2026 00:00:00 GMT</pubDate><language>en-us</language><category>Founding Essay</category><category>About AITi</category></item></channel></rss>