<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:dc="http://purl.org/dc/elements/1.1/">
  <channel>
    <title>Ribbsaeter Systems · Sovereign Systems Journal</title>
    <link>https://www.ribbsaetersystems.com/journal</link>
    <description>Architectural specifications, systems research, and compiler source-levels by Ribbsaeter Systems.</description>
    <language>en-us</language>
    <lastBuildDate>Sun, 22 Nov 2026 15:05:29 GMT</lastBuildDate>
    <atom:link href="https://www.ribbsaetersystems.com/feed.xml" rel="self" type="application/rss+xml"/>
    <item>
      <title>The Convergence on Deterministic AST Invariants: Why Frontier AI Had to Abandon Probabilistic Code Editing</title>
      <link>https://www.ribbsaetersystems.com/journal/ast-invariants-compiler-revolution</link>
      <guid isPermaLink="true">https://www.ribbsaetersystems.com/journal/ast-invariants-compiler-revolution</guid>
      <pubDate>Thu, 29 Oct 2026 23:00:00 GMT</pubDate>
      <description>Between 2024 and October 2026, agentic AI code editing relied on probabilistic string matching and fuzzy diffs. A first-principles systems monograph documenting the mathematical ceiling of probabilistic diffing, Google Antigravity v2.6.0 AST gates, Gemini 4 Carbon benchmarks, and the industry convergence on deterministic compiler verification.</description>
      <category>COMPILER INVARIANTS &amp; FRONTIER AI</category>
      <dc:creator><![CDATA[Patrick Ribbsaeter]]></dc:creator>
    </item>
    <item>
      <title>THE MEMORY WALL: Frontier LLM Inference, Disaggregated Systems &amp; Silicon Hardware Economics</title>
      <link>https://www.ribbsaetersystems.com/journal/the-memory-wall-frontier-llm-inference-disaggregated-systems</link>
      <guid isPermaLink="true">https://www.ribbsaetersystems.com/journal/the-memory-wall-frontier-llm-inference-disaggregated-systems</guid>
      <pubDate>Sat, 24 Oct 2026 00:00:00 GMT</pubDate>
      <description>As frontier AI models scale context windows beyond 128k into multi-million tokens, GPU clusters face an unforgiving physical reality: the Key-Value Cache Memory Wall. A microarchitectural investigation into Disaggregated Prefill &amp; Decode (PD Disaggregation), roofline limits, and $1.25M/year cluster economics.</description>
      <category>FRONTIER AI &amp; SILICON ECONOMICS</category>
      <dc:creator><![CDATA[Patrick Ribbsaeter]]></dc:creator>
    </item>
    <item>
      <title>Recursive Harnessing vs Weight Fine-Tuning: Deconstructing Google Research&apos;s Harness² Architecture in Open-World Agent Runtimes</title>
      <link>https://www.ribbsaetersystems.com/journal/recursive-harnessing-vs-weight-fine-tuning-google-research-harness2</link>
      <guid isPermaLink="true">https://www.ribbsaetersystems.com/journal/recursive-harnessing-vs-weight-fine-tuning-google-research-harness2</guid>
      <pubDate>Sun, 18 Oct 2026 00:00:00 GMT</pubDate>
      <description>Google Research introduces Harness²: a recursive agent harnessing framework that refines execution surfaces on frozen foundation weights. An empirical teardown of task-level Propose–Probe–Compose, lifelong harnessing-level reflection, and why runtime harness synthesis outperforms multi-million-dollar weight fine-tuning across open-world benchmarks.</description>
      <category>RECURSIVE AGENT ARCHITECTURE</category>
      <dc:creator><![CDATA[Patrick Ribbsaeter]]></dc:creator>
    </item>
    <item>
      <title>Detection Without Conformance: Why Cloud AI Scanners Exclude Fix Quality—and the Looming Crisis of Sovereign Code Exfiltration</title>
      <link>https://www.ribbsaetersystems.com/journal/detection-without-conformance-the-sandbox-benchmark-illusion</link>
      <guid isPermaLink="true">https://www.ribbsaetersystems.com/journal/detection-without-conformance-the-sandbox-benchmark-illusion</guid>
      <pubDate>Sun, 18 Oct 2026 00:20:18 GMT</pubDate>
      <description>Commercial AI security benchmarks boast 90% detection rates while quietly excluding fix quality. Simultaneously, cloud sandboxes expose proprietary enterprise IP. Why tier-one engineering teams are moving to local AST invariant gates.</description>
      <category>SOVEREIGN CODE SECURITY</category>
      <dc:creator><![CDATA[Patrick Ribbsaeter]]></dc:creator>
    </item>
    <item>
      <title>The KV Cache Memory Wall: How Multi-Gigabyte Context Windows Throttle GPU Concurrency on H100 Clusters</title>
      <link>https://www.ribbsaetersystems.com/journal/the-kv-cache-memory-wall-and-disaggregated-prefill</link>
      <guid isPermaLink="true">https://www.ribbsaetersystems.com/journal/the-kv-cache-memory-wall-and-disaggregated-prefill</guid>
      <pubDate>Sun, 18 Oct 2026 00:20:15 GMT</pubDate>
      <description>As frontier AI models scale context windows beyond 128k tokens, enterprise GPU clusters face a devastating physical bottleneck: KV cache memory explosion. An architectural investigation into Disaggregated Prefill and Decode (PD Disaggregation).</description>
      <category>FRONTIER AI ARCHITECTURE</category>
      <dc:creator><![CDATA[Patrick Ribbsaeter]]></dc:creator>
    </item>
    <item>
      <title>The Failure of Probabilistic AI Code Review: Why Enterprise CI Demands Compiler-Grade Invariants</title>
      <link>https://www.ribbsaetersystems.com/journal/the-failure-of-probabilistic-ai-code-review</link>
      <guid isPermaLink="true">https://www.ribbsaetersystems.com/journal/the-failure-of-probabilistic-ai-code-review</guid>
      <pubDate>Sun, 18 Oct 2026 00:20:13 GMT</pubDate>
      <description>Over 80% of pull requests generated by AI agents introduce subtle structural defects—including stealth test deletions, hallucinated package dependencies, and CWE regressions. This investigation dissects why probabilistic token predictors cannot enforce program correctness.</description>
      <category>COMPILER INVARIANTS</category>
      <dc:creator><![CDATA[Patrick Ribbsaeter]]></dc:creator>
    </item>
    <item>
      <title>Why I Refused to Build Another LLM Code Reviewer: A Founder&apos;s Notes on Software Invariants</title>
      <link>https://www.ribbsaetersystems.com/journal/why-i-refused-to-build-another-llm-code-reviewer</link>
      <guid isPermaLink="true">https://www.ribbsaetersystems.com/journal/why-i-refused-to-build-another-llm-code-reviewer</guid>
      <pubDate>Sun, 18 Oct 2026 00:20:12 GMT</pubDate>
      <description>When autonomous coding agents began entering production workflows, the software industry assumed secondary generative models could review their code. Here is why I refused to build another chatbot wrapper—and what inspecting thousands of agent PRs revealed.</description>
      <category>FOUNDER ARCHITECTURE</category>
      <dc:creator><![CDATA[Patrick Ribbsaeter]]></dc:creator>
    </item>
  </channel>
</rss>