Journal Archive — Page 17 of 165
2026-09-19
Four Models, One Matchbook: Multi-Model Consensus in the Age of AI
When four frontier AI models independently brainstorm 30 approaches to a technical book, all four converge on a pocket-sized matchbook of verifiable rules.
2026-09-19
Dual-Model Failover: Engineering Independent Backoff Clocks for LLM APIs
Learn how to build resilient dual-model failover with independent backoff clocks for LLM APIs to handle rate limits and availability crashes smoothly.
2026-09-19
Playing Pinball in the Age of AI: A Replayable Workflow Guide
Discover how treating LLMs like stateless pinball tables and using checkable receipts transforms AI coding from vibe-based chaos into reproducible software...
2026-09-19
The Color Blue and the Pipe: Inspecting Code Before Execution
Discover how inspect-before-execute terminal workflows, Unix pipes, and content negotiation bridge the gap between human trust and automated software...
2026-09-19
Edna Mode Rules for AI Workflows: Why Practical Tooling Beats Fragile Abstractions
Discover how practical local-first tooling, verifiable receipts, and zero-cape engineering replace fragile cloud abstractions in modern AI workflows.
2026-09-18
The One-Door Threshold: Engineering Replayable AI Workflows and Quiet Shells
Discover how replacing numeric startup menus with a reproducible command vocabulary creates verifiable, replayable AI workflows in the Age of AI.
2026-09-18
The First Ten Minutes: Engineering Reproducible AI Workflows in the Age of AI
Learn how to build reproducible, checkable AI workflows in ten minutes using Unix pipes, NeoVim, and local-first context compilation without vendor lock-in.
2026-09-17
The Retrieval Cue: Memory Consolidation and Selective Hydration in the Age of AI
Explore how automated codebase curation mirrors biological sleep, moving rich mechanisms into verifiable storage while retaining precise context cues.
2026-09-17
The Voice Asks First: Engineering Verifiable Consent and Silent Defaults
Why automated tools must ask before speaking. How separating the consent ceremony from audio enforcement creates checkable, silent-by-default workflows.
2026-09-17
The Workbench Pin and the Help Screen Census: Verifiable Receipts in the Age of AI
Pinning active build specs to a Jekyll home page reveals how plain text compilers and repeatable probe receipts catch silent tooling bugs before release.