<?xml version="1.0" encoding="UTF-8"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Ostrium Labs — Writing</title><link>https://ostriumlabs.org/</link><atom:link href="https://ostriumlabs.org/feed.xml" rel="self" type="application/rss+xml"/><description>Essays, research notes and unregistered directions from Ostrium Labs. Editorial, not registered research.</description><language>en</language><item><title>The Ostrium Thesis</title><link>https://ostriumlabs.org/writing/the-ostrium-thesis/</link><guid isPermaLink="true">https://ostriumlabs.org/writing/the-ostrium-thesis/</guid><pubDate>Sun, 04 Oct 2026 00:00:00 GMT</pubDate><description>Why time-to-capability matters, how the bottleneck moves, and what increasingly improvable research systems might make accessible. (FOUNDING ESSAY / NON-CANONICAL / NOT A REGISTERED HYPOTHESIS)</description></item><item><title>The Civilization Clock</title><link>https://ostriumlabs.org/writing/civilization-clock/</link><guid isPermaLink="true">https://ostriumlabs.org/writing/civilization-clock/</guid><pubDate>Sat, 03 Oct 2026 00:00:00 GMT</pubDate><description>A day-sized illustration of selected technological history, and the limits of looking backward to imagine what comes next. (NON-CANONICAL ESSAY / NOT REGISTERED / ILLUSTRATIVE METAPHOR)</description></item><item><title>The Experience Envelope</title><link>https://ostriumlabs.org/writing/experience-envelope/</link><guid isPermaLink="true">https://ostriumlabs.org/writing/experience-envelope/</guid><pubDate>Sat, 03 Oct 2026 00:00:00 GMT</pubDate><description>How much previously inaccessible reality can become something we perceive, manipulate, inhabit, or experience? (NON-CANONICAL ESSAY / NOT REGISTERED / DIRECTIONAL AMBITION)</description></item><item><title>Loams: Design Invariants for AI-Native Storage</title><link>https://ostriumlabs.org/writing/loams-architectural-invariants/</link><guid isPermaLink="true">https://ostriumlabs.org/writing/loams-architectural-invariants/</guid><pubDate>Sat, 03 Oct 2026 00:00:00 GMT</pubDate><description>Why we built object-storage-first durable retrieval, and why applications must be able to leave without rewriting their code. (NON-CANONICAL / NOT REGISTERED / RESEARCH NOTE — Technical System Specification. Grounded in Ledger Entry #001.)</description></item><item><title>Notes on Epistemic Discipline in Frontier Systems</title><link>https://ostriumlabs.org/writing/notes-on-epistemic-discipline/</link><guid isPermaLink="true">https://ostriumlabs.org/writing/notes-on-epistemic-discipline/</guid><pubDate>Sat, 03 Oct 2026 00:00:00 GMT</pubDate><description>Why benchmarks, prototypes, and reusable artifacts require different evidence, and how the Capability Ledger distinguishes them. (NON-CANONICAL / NOT REGISTERED / FIELD NOTE — Exploratory Observation. Does not constitute verified evidence.)</description></item><item><title>The Latency of Reality</title><link>https://ostriumlabs.org/writing/the-latency-of-reality/</link><guid isPermaLink="true">https://ostriumlabs.org/writing/the-latency-of-reality/</guid><pubDate>Sat, 03 Oct 2026 00:00:00 GMT</pubDate><description>Why the interval between a question and a validated capability matters, and which parts of that interval might be compressible. (NON-CANONICAL / NOT REGISTERED / ESSAY — Foundational Philosophical Analysis. Not a pre-registered hypothesis.)</description></item><item><title>Recursive Compression</title><link>https://ostriumlabs.org/writing/recursive-compression/</link><guid isPermaLink="true">https://ostriumlabs.org/writing/recursive-compression/</guid><pubDate>Sat, 03 Oct 2026 00:00:00 GMT</pubDate><description>What changes when intelligence improves the tools—and eventually the methods—used to improve intelligence? (UNREGISTERED FIELD NOTE / NON-CANONICAL / NO VALIDATION CLAIM)</description></item></channel></rss>