A weekly long-form post on what's actually happening in AI adoption — the questions teams are asking, the failure modes worth naming, and the patterns showing up across industries. Less hype, more pragmatism.
Most pieces first appear as a video and long-form post on LinkedIn, then land here for the archive.
It's happening today. This morning. Maybe an hour ago. The risk isn't AI. The risk is not knowing. Most of the value of an AI policy sits in answering three questions clearly enough that the person on your team using ChatGPT this afternoon knows what's allowed.
In 2023, Samsung engineers pasted proprietary source code into ChatGPT. The trade secrets went with it. Most coverage missed the actual lesson: the problem wasn't that engineers used AI. The problem was that nobody had told them what was OK to put into it.
It did it in an afternoon. I drank coffee. A competitive landscape analysis that would traditionally cost six figures and three weeks — done on my laptop, with the AI subscription I was already paying for. The work that used to flow to outside consultants is staying inside.
That sentence is a contradiction, and for most of my career it wouldn't have been possible. It's possible now because of a quiet shift almost nobody outside of tech is talking about: vibe coding. The most underrated productivity move available to non-technical professionals right now.
Most AI conversations focus on big projects. The everyday value comes from a few simple, repeatable prompts I use almost every working day — for summarizing, pressure-testing decisions, and drafting difficult messages. None of them clever. All of them save real time.
A room of eight people. Three priorities that don't quite line up. Forty minutes of discussion, ten minutes of agreement, a notebook full of half-sentences. The traditional move is to stare at your notes until they organize themselves. They never do. The new move is faster.
When industrial leaders tell me they're behind on AI, I ask them to walk me through what their team did last week. It almost always turns out they're using AI — they just didn't know they were. Predictive maintenance, route optimization, demand forecasting, safety monitoring. Read the list.
Why this exists, what to expect, and what won't be here. A short intro to the cadence and the kind of writing you'll find — practical observations from running AI workshops, drafting policies with real teams, and getting things wrong often enough to have opinions about it.
Want these in your inbox? Email subscription coming soon. For now, follow on LinkedIn — that's where each week's piece lands first. Or subscribe via RSS.