corpus.blog
56,966 blogs · [ { "id": "01a0d50c-e565-73c9-a554-2952fd4e0fad", "title": "The Second Harness Tax", "url": "https://ai.rundatarun.io/ai-development-agents/the-second-harness-tax", "published_at": "2026-09-22T12:00:00+00:00" }, { "id": "01a0d50c-e565-73c9-a554-2952fda28605", "title": "Four Upgrades to the Agent Loop. None of Them Moved the Number.", "url": "https://ai.rundatarun.io/ai-development-agents/four-upgrades-none-moved-the-number", "published_at": "2026-09-14T12:00:00+00:00" }, { "id": "01a0d50c-e565-73c9-a554-2952fe209007", "title": "Claude Code Is Getting Mods. Here Is What That Actually Changes.", "url": "https://ai.rundatarun.io/ai-development-agents/claude-code-is-getting-mods", "published_at": "2026-09-14T12:00:00+00:00" }, { "id": "01a0d50c-e565-73c9-a554-2952feda699f", "title": "62.7 and 99.9 Are the Same Model", "url": "https://ai.rundatarun.io/ai-development-agents/62-7-and-99-9-are-the-same-model", "published_at": "2026-09-12T12:00:00+00:00" }, { "id": "01a08721-af12-7376-8fd7-0bc75404e0e6", "title": "I checked the 90% token cut against 30 days of my own Claude Code. The hook would have fired on 3.5% of it.", "url": "https://ai.rundatarun.io/ai-development-agents/i-checked-the-90-percent-token-cut-on-my-own-30-days", "published_at": "2026-09-06T12:00:00+00:00" }, { "id": "01a08721-af12-7376-8fd7-0bc754d5b601", "title": "Everything In The Family Holds The Key", "url": "https://ai.rundatarun.io/ai-systems-architecture/stealing-reasoning-traces", "published_at": "2026-08-12T12:00:00+00:00" }, { "id": "01a08721-af12-7376-8fd7-0bc754e225eb", "title": "A Loop With Better Marketing", "url": "https://ai.rundatarun.io/ai-development-agents/a-loop-with-better-marketing", "published_at": "2026-08-05T12:00:00+00:00" }, { "id": "01a08721-af12-7376-8fd7-0bc75570b8f2", "title": "Four Slots Left", "url": "https://ai.rundatarun.io/ai-development-agents/four-slots-left", "published_at": "2026-08-04T12:00:00+00:00" }, { "id": "01a08721-af12-7376-8fd7-0bc7562a8202", "title": "Tabular foundation models got acquired before they got benchmarked", "url": "https://ai.rundatarun.io/cutting-edge-ai/tabular-foundation-models-acquired-before-benchmarked", "published_at": "2026-07-27T12:00:00+00:00" }, { "id": "01a08721-af12-7376-8fd7-0bc75636cba8", "title": "Leading From the Other Side (the Month-Three Checkpoint)", "url": "https://ai.rundatarun.io/ai-development-agents/field-guide-09-leading-from-the-other-side", "published_at": "2026-07-20T12:00:00+00:00" }, { "id": "01a08721-af12-7376-8fd7-0bc756efdbe0", "title": "From One Operator to a Team (the Phase Everyone Skips)", "url": "https://ai.rundatarun.io/ai-development-agents/field-guide-08-one-operator-to-a-team", "published_at": "2026-07-18T12:00:00+00:00" }, { "id": "01a08721-af12-7376-8fd7-0bc7575bea3f", "title": "Build a free Audible replacement in an afternoon with Claude Code", "url": "https://ai.rundatarun.io/practical-applications/free-audible-replacement-claude-code", "published_at": "2026-07-17T12:00:00+00:00" }, { "id": "01a08721-af12-7376-8fd7-0bc75769f9b3", "title": "Build the Harness Out (Weeks 2 to 6)", "url": "https://ai.rundatarun.io/ai-development-agents/field-guide-07-build-the-harness-out", "published_at": "2026-07-16T12:00:00+00:00" }, { "id": "01a08721-af12-7376-8fd7-0bc75788965e", "title": "Bonsai 27B: Frontier Reasoning at 1.7 Bits a Weight", "url": "https://ai.rundatarun.io/cutting-edge-ai/bonsai-27b-native-low-bit", "published_at": "2026-07-15T12:00:00+00:00" }, { "id": "01a08721-af12-7376-8fd7-0bc75866e8da", "title": "Three Sessions, One Company: What Parallel AI Sessions Actually Cost", "url": "https://ai.rundatarun.io/ai-development-agents/three-sessions-one-company", "published_at": "2026-07-14T12:00:00+00:00" }, { "id": "01a08721-af12-7376-8fd7-0bc7588f5a0e", "title": "Your First Saturday With Claude Code", "url": "https://ai.rundatarun.io/ai-development-agents/field-guide-06-your-first-saturday", "published_at": "2026-07-14T12:00:00+00:00" }, { "id": "01a08721-af12-7376-8fd7-0bc75918d570", "title": "The Five Leadership Primitives Already Transfer", "url": "https://ai.rundatarun.io/ai-development-agents/field-guide-05-five-primitives-transfer", "published_at": "2026-07-12T12:00:00+00:00" }, { "id": "01a08721-af12-7376-8fd7-0bc759911868", "title": "Inventory Your Harness: The Six Components", "url": "https://ai.rundatarun.io/ai-development-agents/field-guide-04-inventory-your-harness", "published_at": "2026-07-10T12:00:00+00:00" }, { "id": "01a08721-af12-7376-8fd7-0bc75a4cac18", "title": "Maximizing Toward the Local Minimum: How Fast Optimization Drifts, and the Human Habit That Catches It", "url": "https://ai.rundatarun.io/practical-applications/maximizing-toward-the-local-minimum", "published_at": "2026-07-08T12:00:00+00:00" }, { "id": "01a08721-af12-7376-8fd7-0bc75adbbd86", "title": "Why Your Org Can't Cross (Run Your Own Pilot Numbers)", "url": "https://ai.rundatarun.io/ai-development-agents/field-guide-03-run-the-pilot-numbers", "published_at": "2026-07-08T12:00:00+00:00" }, { "id": "01a08721-af12-7376-8fd7-0bc75bd21410", "title": "Hear the Three Tiers (and Name the Failure Mode)", "url": "https://ai.rundatarun.io/ai-development-agents/field-guide-02-hear-the-three-tiers", "published_at": "2026-07-06T12:00:00+00:00" }, { "id": "01a08721-af12-7376-8fd7-0bc75bd2e91e", "title": "Which Side of the Build Gap Are You On? (Run the Scorecard)", "url": "https://ai.rundatarun.io/ai-development-agents/field-guide-01-which-side-of-the-gap", "published_at": "2026-07-04T12:00:00+00:00" }, { "id": "01a08721-af12-7376-8fd7-0bc75c53525c", "title": "The Harness That Wrote the Book", "url": "https://ai.rundatarun.io/ai-development-agents/the-harness-that-wrote-the-book", "published_at": "2026-07-01T12:00:00+00:00" }, { "id": "01a08721-af12-7376-8fd7-0bc75cb9cf29", "title": "Keeping the Node Smoking: What Eight H100s Buy, and the Engine That Keeps Them Full", "url": "https://ai.rundatarun.io/practical-applications/keeping-the-node-smoking", "published_at": "2026-06-30T12:00:00+00:00" }, { "id": "01a08721-af12-7376-8fd7-0bc75d93a02d", "title": "Sol vs Terra vs Luna: A Builder's Guide to GPT-5.6", "url": "https://ai.rundatarun.io/cutting-edge-ai/gpt-5-6-sol-terra-luna-builders-read", "published_at": "2026-06-27T12:00:00+00:00" }, { "id": "01a08721-af12-7376-8fd7-0bc75de2d156", "title": "Waking the Research Engine: Three Walls, One Experiment", "url": "https://ai.rundatarun.io/practical-applications/waking-the-research-engine", "published_at": "2026-06-26T12:00:00+00:00" }, { "id": "01a08721-af12-7376-8fd7-0bc75eaa3e26", "title": "Day One: Standing Up the Inference Platform", "url": "https://ai.rundatarun.io/practical-applications/day-one-standing-up-the-inference-platform", "published_at": "2026-06-25T12:00:00+00:00" }, { "id": "01a08721-af12-7376-8fd7-0bc75f20d6f8", "title": "From a DGX Spark to a Borrowed Node: A Retinal-AI Lab", "url": "https://ai.rundatarun.io/practical-applications/from-a-spark-to-a-borrowed-cluster", "published_at": "2026-06-24T12:00:00+00:00" }, { "id": "01a08721-af12-7376-8fd7-0bc75fa51466", "title": "Two Camps of Reasoning Post-Training: Checkable or Not", "url": "https://ai.rundatarun.io/cutting-edge-ai/two-camps-reasoning-post-training", "published_at": "2026-06-24T12:00:00+00:00" }, { "id": "01a08721-af12-7376-8fd7-0bc7601bbc46", "title": "The Composition Threshold: Bigger Library, Better Agent?", "url": "https://ai.rundatarun.io/ai-systems-architecture/the-composition-threshold", "published_at": "2026-06-21T12:00:00+00:00" }, { "id": "01a08721-af12-7376-8fd7-0bc7604bf0f7", "title": "Two Bets on Where Agent Context Lives", "url": "https://ai.rundatarun.io/ai-systems-architecture/two-bets-on-where-agent-context-lives", "published_at": "2026-06-17T12:00:00+00:00" }, { "id": "01a08721-af12-7376-8fd7-0bc76070da19", "title": "Build an LLM Triage Layer That Can't Freelance", "url": "https://ai.rundatarun.io/ai-systems-architecture/grounded-llm-triage-layer", "published_at": "2026-06-11T12:00:00+00:00" }, { "id": "01a08721-af12-7376-8fd7-0bc7610494ee", "title": "The Control-Plane Bet: Code-as-Action in Search and Agents", "url": "https://ai.rundatarun.io/ai-systems-architecture/the-control-plane-bet-code-as-action", "published_at": "2026-06-02T12:00:00+00:00" }, { "id": "01a08721-af12-7376-8fd7-0bc76165c52f", "title": "Mneme: Semantic Recall for Your Claude Code Sessions", "url": "https://ai.rundatarun.io/practical-applications/mneme-semantic-recall", "published_at": "2026-05-29T12:00:00+00:00" }, { "id": "01a08721-af12-7376-8fd7-0bc76169fa11", "title": "Managing a Homelab's LLM API Keys With LiteLLM", "url": "https://ai.rundatarun.io/practical-applications/litellm-keys-easy-hunting-them-down-hard", "published_at": "2026-05-21T12:00:00+00:00" }, { "id": "01a08721-af12-7376-8fd7-0bc761c1a383", "title": "Standing Up a Homelab URL Shortener With Shlink", "url": "https://ai.rundatarun.io/practical-applications/url-shortener-homelab-shlink-nymio", "published_at": "2026-05-19T12:00:00+00:00" }, { "id": "01a08721-af12-7376-8fd7-0bc7622c62ea", "title": "Building Claude Code Skills by Conversation: Dossier Pattern", "url": "https://ai.rundatarun.io/ai-development-agents/building-claude-code-skills-dossier-pattern", "published_at": "2026-05-18T12:00:00+00:00" }, { "id": "01a08721-af12-7376-8fd7-0bc7628df546", "title": "Defending Your Homelab Fleet From npm Supply-Chain Attacks", "url": "https://ai.rundatarun.io/practical-applications/defending-homelab-npm-supply-chain", "published_at": "2026-05-13T12:00:00+00:00" }, { "id": "01a08721-af12-7376-8fd7-0bc763327df7", "title": "Codex vs Claude Code vs OpenCode: From Inside the Harness", "url": "https://ai.rundatarun.io/ai-development-agents/codex-vs-claude-code-vs-opencode", "published_at": "2026-05-12T12:00:00+00:00" }, { "id": "01a08721-af12-7376-8fd7-0bc76368959c", "title": "Anthropic Said 'Feels Infinite.' Dreaming Is Where It Breaks", "url": "https://ai.rundatarun.io/ai-development-agents/anthropic-dreaming-continual-learning", "published_at": "2026-05-08T12:00:00+00:00" }, { "id": "01a08721-af12-7376-8fd7-0bc764499c19", "title": "RecursiveMAS and the End of Text-Mediated Agent Systems", "url": "https://ai.rundatarun.io/ai-development-agents/recursive-mas-latent-space-agent-communication", "published_at": "2026-05-04T12:00:00+00:00" }, { "id": "01a08721-af12-7376-8fd7-0bc764b247d8", "title": "Teaching Gemma 4 E2B to Draft Like DFlash: A Build Log", "url": "https://ai.rundatarun.io/cutting-edge-ai/gemma4-diffusion-drafter-build-log", "published_at": "2026-05-01T12:00:00+00:00" }, { "id": "01a08721-af12-7376-8fd7-0bc765442f32", "title": "Reproducing Warp Decode: MoE Parallelism on Blackwell", "url": "https://ai.rundatarun.io/ai-systems-architecture/reproducing-warp-decode-blackwell", "published_at": "2026-04-09T12:00:00+00:00" }, { "id": "01a08721-af12-7376-8fd7-0bc765d1b44b", "title": "TurboQuant: 1,000 Experiments, One Uncomfortable Finding", "url": "https://ai.rundatarun.io/ai-development-agents/turboquant-kv-quantization", "published_at": "2026-03-26T12:00:00+00:00" }, { "id": "01a08721-af12-7376-8fd7-0bc766019240", "title": "AutoResearch on Blackwell GB10: 151 Experiments Overnight", "url": "https://ai.rundatarun.io/practical-applications/autoresearch-blackwell-gb10-151-experiments", "published_at": "2026-03-14T12:00:00+00:00" }, { "id": "01a08721-af12-7376-8fd7-0bc766021c7a", "title": "Control Your Smart Home From the Terminal", "url": "https://ai.rundatarun.io/practical-applications/controlling-your-home-from-the-terminal", "published_at": "2026-02-28T12:00:00+00:00" }, { "id": "01a08721-af12-7376-8fd7-0bc766dfae4d", "title": "AI as Exoskeleton, Not Coworker: Human-AI System Patterns", "url": "https://ai.rundatarun.io/emerging-trends/ai-as-exoskeleton-not-coworker", "published_at": "2026-02-26T12:00:00+00:00" }, { "id": "01a08721-af12-7376-8fd7-0bc7670cf72a", "title": "The Model IS the Computer: Compute-in-Memory for AI", "url": "https://ai.rundatarun.io/cutting-edge-ai/model-is-the-computer-compute-in-memory", "published_at": "2026-02-26T12:00:00+00:00" }, { "id": "01a08721-af12-7376-8fd7-0bc767df9492", "title": "Fix macOS Window Chaos With Hammerspoon and Karabiner", "url": "https://ai.rundatarun.io/practical-applications/fixing-macos-window-chaos-hammerspoon-karabiner", "published_at": "2026-02-20T12:00:00+00:00" }, { "id": "01a08721-af12-7376-8fd7-0bc7688d3572", "title": "I Built an Autonomous AI Agent Squad for $10/Month", "url": "https://ai.rundatarun.io/ai-development-agents/autonomous-ai-agent-squad-10-dollars-month", "published_at": "2026-02-06T12:00:00+00:00" }, { "id": "01a08721-af12-7376-8fd7-0bc7696fb814", "title": "When ARIA Crashed the DGX: GPU Monitoring in 5 Minutes", "url": "https://ai.rundatarun.io/practical-applications/when-aria-crashed-the-dgx", "published_at": "2026-01-17T12:00:00+00:00" }, { "id": "01a08721-af12-7376-8fd7-0bc76a6ac010", "title": "When AI Knows What You Mean, Not Just What You Say", "url": "https://ai.rundatarun.io/ai-development-agents/ambient-intelligence-intent-over-instructions", "published_at": "2026-01-16T12:00:00+00:00" }, { "id": "01a08721-af13-7287-bef1-2311a912e2bf", "title": "Hardening a Production System with 40 Parallel AI Agents", "url": "https://ai.rundatarun.io/practical-applications/concentric-system-analysis-with-claude", "published_at": "2026-01-15T12:00:00+00:00" }, { "id": "01a08721-af13-7287-bef1-2311a93d301e", "title": "MCPHub: Managing MCP Servers Across Multiple AI Tools", "url": "https://ai.rundatarun.io/practical-applications/mcphub-managing-mcp-servers-across-ai-tools", "published_at": "2026-01-14T12:00:00+00:00" } ] posts Claim your blog
Back to ai.rundatarun.io
Blog · corpus.blog/blogs/ai.rundatarun.io/posts

ai.rundatarun.io

ai.rundatarun.io

2026

The Second Harness Taxoriginal ↗
22 Sept 2026
Four Upgrades to the Agent Loop. None of Them Moved the Number.original ↗
14 Sept 2026
Claude Code Is Getting Mods. Here Is What That Actually Changes.original ↗
14 Sept 2026
62.7 and 99.9 Are the Same Modeloriginal ↗
12 Sept 2026
I checked the 90% token cut against 30 days of my own Claude Code. The hook would have fired on 3.5% of it.original ↗
6 Sept 2026
Everything In The Family Holds The Keyoriginal ↗
12 Aug 2026
A Loop With Better Marketingoriginal ↗
5 Aug 2026
Four Slots Leftoriginal ↗
4 Aug 2026
Tabular foundation models got acquired before they got benchmarkedoriginal ↗
27 Jul 2026
Leading From the Other Side (the Month-Three Checkpoint)original ↗
20 Jul 2026
From One Operator to a Team (the Phase Everyone Skips)original ↗
18 Jul 2026
Build a free Audible replacement in an afternoon with Claude Codeoriginal ↗
17 Jul 2026
Build the Harness Out (Weeks 2 to 6)original ↗
16 Jul 2026
Bonsai 27B: Frontier Reasoning at 1.7 Bits a Weightoriginal ↗
15 Jul 2026
Three Sessions, One Company: What Parallel AI Sessions Actually Costoriginal ↗
14 Jul 2026
Your First Saturday With Claude Codeoriginal ↗
14 Jul 2026
The Five Leadership Primitives Already Transferoriginal ↗
12 Jul 2026
Inventory Your Harness: The Six Componentsoriginal ↗
10 Jul 2026
Maximizing Toward the Local Minimum: How Fast Optimization Drifts, and the Human Habit That Catches Itoriginal ↗
8 Jul 2026
Why Your Org Can't Cross (Run Your Own Pilot Numbers)original ↗
8 Jul 2026
Hear the Three Tiers (and Name the Failure Mode)original ↗
6 Jul 2026
Which Side of the Build Gap Are You On? (Run the Scorecard)original ↗
4 Jul 2026
The Harness That Wrote the Bookoriginal ↗
1 Jul 2026
Keeping the Node Smoking: What Eight H100s Buy, and the Engine That Keeps Them Fulloriginal ↗
30 Jun 2026
Sol vs Terra vs Luna: A Builder's Guide to GPT-5.6original ↗
27 Jun 2026
Waking the Research Engine: Three Walls, One Experimentoriginal ↗
26 Jun 2026
Day One: Standing Up the Inference Platformoriginal ↗
25 Jun 2026
From a DGX Spark to a Borrowed Node: A Retinal-AI Laboriginal ↗
24 Jun 2026
Two Camps of Reasoning Post-Training: Checkable or Notoriginal ↗
24 Jun 2026
The Composition Threshold: Bigger Library, Better Agent?original ↗
21 Jun 2026
Two Bets on Where Agent Context Livesoriginal ↗
17 Jun 2026
Build an LLM Triage Layer That Can't Freelanceoriginal ↗
11 Jun 2026
The Control-Plane Bet: Code-as-Action in Search and Agentsoriginal ↗
2 Jun 2026
Mneme: Semantic Recall for Your Claude Code Sessionsoriginal ↗
29 May 2026
Managing a Homelab's LLM API Keys With LiteLLMoriginal ↗
21 May 2026
Standing Up a Homelab URL Shortener With Shlinkoriginal ↗
19 May 2026
Building Claude Code Skills by Conversation: Dossier Patternoriginal ↗
18 May 2026
Defending Your Homelab Fleet From npm Supply-Chain Attacksoriginal ↗
13 May 2026
Codex vs Claude Code vs OpenCode: From Inside the Harnessoriginal ↗
12 May 2026
Anthropic Said 'Feels Infinite.' Dreaming Is Where It Breaksoriginal ↗
8 May 2026
RecursiveMAS and the End of Text-Mediated Agent Systemsoriginal ↗
4 May 2026
Teaching Gemma 4 E2B to Draft Like DFlash: A Build Logoriginal ↗
1 May 2026
Reproducing Warp Decode: MoE Parallelism on Blackwelloriginal ↗
9 Apr 2026
TurboQuant: 1,000 Experiments, One Uncomfortable Findingoriginal ↗
26 Mar 2026
AutoResearch on Blackwell GB10: 151 Experiments Overnightoriginal ↗
14 Mar 2026
Control Your Smart Home From the Terminaloriginal ↗
28 Feb 2026
AI as Exoskeleton, Not Coworker: Human-AI System Patternsoriginal ↗
26 Feb 2026
The Model IS the Computer: Compute-in-Memory for AIoriginal ↗
26 Feb 2026
Fix macOS Window Chaos With Hammerspoon and Karabineroriginal ↗
20 Feb 2026
I Built an Autonomous AI Agent Squad for $10/Monthoriginal ↗
6 Feb 2026
When ARIA Crashed the DGX: GPU Monitoring in 5 Minutesoriginal ↗
17 Jan 2026
When AI Knows What You Mean, Not Just What You Sayoriginal ↗
16 Jan 2026
Hardening a Production System with 40 Parallel AI Agentsoriginal ↗
15 Jan 2026
MCPHub: Managing MCP Servers Across Multiple AI Toolsoriginal ↗
14 Jan 2026