Builder Log
25 notes on this leg
Builder Log is where I document experiments with agentic AI, automation workflows, local systems, and the messy work of making useful tools reliable. These notes are less about the tool itself and more about what the build teaches.
Shadow AI in Small Organizations: A 30-Minute Discovery Checklist
Pax8's survey suggests AI building is outpacing review. A practical discovery checklist for small-business, school, and nonprofit IT leaders.
Actually-Free AI Learning: An October Check Before November
Two verified free AI learning paths, plus a warning about course pages where free enrollment does not prove the full…
What the FSB’s AI Sound Practices Mean for Human Oversight
The Financial Stability Board's October 2026 AI guidance is not a binding rulebook. Its risk-based oversight test is still useful…
Salesforce Just Gave Its AI Agents Names and Job Titles — Here’s What That Actually Changes
Salesforce's Agentforce 360 gives seven AI agents names and job titles, plus a governance control plane. Here's what's genuinely new…
When You Tell an AI to Stop, What Actually Stops?
A quiet voice assistant may still be working. Learn the difference between interruption, cancellation, and undo—and what to check before…
OpenAI’s 10,000-Agent Math Run: Why a Formal Proof Still Needs Independent Review
OpenAI has released a claimed Navier–Stokes solution and Lean formalization. Here is how to separate agent scale, proof checking and…
ChatGPT Images 2.5: A Seven-Image Gallery of Roads, Agents and Ideas
Seven GPT-Image-2.5 examples inspired by riding, AI agents and useful workflows—with exact prompts and a controlled before-and-after edit.
Human in the Loop, on the Loop, or over the Loop? A Plain-English AI Oversight Guide
Approval, monitoring, and governance are different jobs. Use this practical comparison to assign human oversight to the actual risk of…
AI Agent Productivity Metrics: What OpenAI’s 3.1 Agent-Workdays Actually Tell Us
OpenAI’s agent-runtime figures show growing automation capacity. Here is a practical scorecard for measuring accepted work, review effort, and actual…
Where Should Your AI Agent Actually Live? A Plain-English Guide to Choosing an AI Agent Hub
Where Should Your AI Agent Actually Live? A Plain-English Guide to Choosing an AI Agent Hub For about two years,…
