<feed xmlns="http://www.w3.org/2005/Atom"> <id>https://jetxu-llm.github.io/</id><title>Jet Xu | DocMason</title><subtitle>Jet Xu writes about DocMason, local-first knowledge bases, and the systems needed to turn messy private files into reliable, evidence-backed context for serious work.</subtitle> <updated>2026-09-15T22:59:23+08:00</updated> <author> <name>Jet Xu</name> <uri>https://jetxu-llm.github.io/</uri> </author><link rel="self" type="application/atom+xml" href="https://jetxu-llm.github.io/feed.xml"/><link rel="alternate" type="text/html" hreflang="en" href="https://jetxu-llm.github.io/"/> <generator uri="https://jekyllrb.com/" version="4.4.1">Jekyll</generator> <rights> © 2026 Jet Xu </rights> <icon>/assets/img/favicons/favicon.ico</icon> <logo>/assets/img/favicons/favicon-96x96.png</logo> <entry><title>AI-Native Is an Inversion of Control</title><link href="https://jetxu-llm.github.io/posts/ai-native-is-an-inversion-of-control/" rel="alternate" type="text/html" title="AI-Native Is an Inversion of Control" /><published>2026-09-15T08:00:00+08:00</published> <updated>2026-09-15T08:00:00+08:00</updated> <id>https://jetxu-llm.github.io/posts/ai-native-is-an-inversion-of-control/</id> <content type="text/html" src="https://jetxu-llm.github.io/posts/ai-native-is-an-inversion-of-control/" /> <author> <name>Jet Xu</name> </author> <category term="AI" /> <category term="Architecture" /> <summary>The difference between AI-assisted and AI-native work is who keeps the process moving. Agents run the case; people set its purpose, limits, and terms of approval.</summary> </entry> <entry><title>AI Slop Is a Situational Awareness Problem</title><link href="https://jetxu-llm.github.io/posts/ai-slop-is-a-situational-awareness-problem/" rel="alternate" type="text/html" title="AI Slop Is a Situational Awareness Problem" /><published>2026-07-05T00:00:00+08:00</published> <updated>2026-07-05T00:00:00+08:00</updated> <id>https://jetxu-llm.github.io/posts/ai-slop-is-a-situational-awareness-problem/</id> <content type="text/html" src="https://jetxu-llm.github.io/posts/ai-slop-is-a-situational-awareness-problem/" /> <author> <name>Jet Xu</name> </author> <category term="AI" /> <category term="Model Evaluation" /> <summary>Two blind copywriting tests for a project called CodexWork showed that AI slop is rarely about bad English. It is a failure to read where the visitor is standing: which page they are on, what they already know, and what the next click should feel like.</summary> </entry> <entry><title>The Future of AI Code Review Is a Digital Staff Engineer</title><link href="https://jetxu-llm.github.io/posts/the-future-of-ai-code-review-is-a-digital-staff-engineer/" rel="alternate" type="text/html" title="The Future of AI Code Review Is a Digital Staff Engineer" /><published>2026-06-26T00:00:00+08:00</published> <updated>2026-06-26T00:00:00+08:00</updated> <id>https://jetxu-llm.github.io/posts/the-future-of-ai-code-review-is-a-digital-staff-engineer/</id> <content type="text/html" src="https://jetxu-llm.github.io/posts/the-future-of-ai-code-review-is-a-digital-staff-engineer/" /> <author> <name>Jet Xu</name> </author> <category term="Engineering" /> <category term="AI" /> <summary>Generic AI reviewers fail where enterprise context begins. The next valuable AI code reviewer will behave like an internal digital staff engineer: platform-aware, company-aware, human-looped, and accountable for merge trust.</summary> </entry> <entry><title>The Death of the God Model: Why True AGI Requires a Split Brain Architecture</title><link href="https://jetxu-llm.github.io/posts/the-death-of-the-god-model/" rel="alternate" type="text/html" title="The Death of the God Model: Why True AGI Requires a Split Brain Architecture" /><published>2026-06-02T22:30:00+08:00</published> <updated>2026-07-15T23:52:08+08:00</updated> <id>https://jetxu-llm.github.io/posts/the-death-of-the-god-model/</id> <content type="text/html" src="https://jetxu-llm.github.io/posts/the-death-of-the-god-model/" /> <author> <name>Jet Xu</name> </author> <category term="AI" /> <category term="Architecture" /> <summary>The AI industry's pursuit of a single omnipotent God Model is a dead end. Due to the mathematical paradoxes of RLHF and alignment, models are inevitably bifurcating into specialized hemispheres—and the future belongs to Agentic Frontal Lobe orchestration.</summary> </entry> <entry><title>I Built repo-graph-rag Before the Code Graph Wave</title><link href="https://jetxu-llm.github.io/posts/i-built-repo-graph-rag-before-the-code-graph-wave/" rel="alternate" type="text/html" title="I Built repo-graph-rag Before the Code Graph Wave" /><published>2026-05-05T20:00:00+08:00</published> <updated>2026-05-05T23:22:12+08:00</updated> <id>https://jetxu-llm.github.io/posts/i-built-repo-graph-rag-before-the-code-graph-wave/</id> <content type="text/html" src="https://jetxu-llm.github.io/posts/i-built-repo-graph-rag-before-the-code-graph-wave/" /> <author> <name>Jet Xu</name> </author> <category term="Engineering" /> <category term="AI" /> <summary>In early 2025 I started repo-graph-rag to build token-free, traversable repository intelligence. The recent code-graph wave validates the problem, but also explains why I moved upstream anyway.</summary> </entry> </feed>
