Artificial intelligence applied to real work: agents, automation, tools and pipelines I use every day to produce more with less friction. No abstract AI talk — I cover Claude Code, Obsidian with semantic search, pipelines that write and publish content, self-correcting agents. What I build, what I break, and what I learn along the way.
Valencia City Council awarded its animal shelters to a pest-control joint venture. I ran an adversarial AI peer review between Claude and Codex. Verdict: net worsening for the animals.
When AI Judges a Shelter Deal: Valencia's Dog Dilemma Leer más »
gbrain beat grep on every relational and semantic query across 31 tests — for under $0.01. See how semantic search in Obsidian actually works.
Semantic Search in Obsidian: gbrain vs grep Tested Leer más »
Manual approval on every piece is the bottleneck killing your output. See how to automate AI content end to end—and scale. Read the full breakdown.
Automate AI Content With Zero Manual Approvals Leer más »
A self-improving agent can learn a lot, but should never touch SOUL.md or AGENTS.md without human approval. Here is why the gate matters.
Why Self-Improving AI Agents Need an Approval Gate Leer más »
Anthropic is quietly testing the removal of Claude Code from its Pro plan. What it means and why platform dependency is a real risk.
Anthropic Tests Removing Claude Code from the Pro Plan Leer más »
AI is not just a threat — it is breaking down barriers. How I learned smart home automation in a month and started building real things with AI.
The AI Revolution Can Also Make Us Better and Happier Leer más »
OpenAI adds native sandbox execution, configurable memory, and file tools to its Agents SDK. What it means for real-world AI agent deployments.
OpenAI Upgrades Agents SDK with Native Sandbox Execution Leer más »