Volume 53: What Would Make AI Progress Matter to You?
AI evaluation starts with a real task and a clear definition of success. Turn a familiar failure into a repeatable test, then check whether a new prompt or tool improves the result.
Volume 52: A Year In and Still Not an AI Expert
A rule against putting confidential data into public AI tools is a prohibition, not a policy. A real AI policy answers five questions about disclosure, editorial control, data leaving the building, training, and what happens when the tool changes.
Volume 51: You Can Only Audit What You Already Understand
The rule that makes you label AI work has three conditions, and most workplace writing fails the first one. A watermark is not a label, an unmarked document proves nothing, and in most jobs the disclosure duty belongs to your employer rather than to you.
Volume 50: The Work You Wrote Off as Impossible
The EU AI Act's AI literacy duty took effect in February 2025 and became enforceable in August 2026. It reaches US companies whose AI use touches people in the EU, and it asks whether the staff using AI tools were told what those tools get wrong.
Volume 49: Inside Every AI Win Is Someone Who Knows the Problem
Before you recommend an AI tool at work, four questions decide if it is actually safe: what it optimizes for, who tested it, what it does not know, and how you would catch it being wrong.
Volume 48: The Failure Mode That Looks Like Great Work
Months before a new AI model reaches you, teams are paid to attack it, and the guardrail that refuses your harmless request is the visible edge of that testing. Plus a one-sentence stop that keeps AI from finishing work it should refuse, and a ten-minute cure for AI news guilt.
Volume 47: The Watermark Cannot Read Intent
An AI agent can hit every number in an instruction and still fail the task, because a content filter blocks what a model says, not what it is trying to do. This piece shows how to write goals a system cannot quietly game.
Volume 46: Audit the Rules Before You Blame the Model
A prompt you re-paste every week belongs in an agent, not your chat history. Microsoft 365 Copilot Agent Builder loads the role, rules, and knowledge sources once, so each week you hand it only what changed, and the same pattern runs in Claude, ChatGPT, and Gemini.