The Day at a Glance

• Built the infrastructure to turn a day’s raw work into a published narrative, automatically, every single day
• The strange experience of using AI to construct the journal that documents using AI
• Caught myself editing differently than I did six months ago, and realized content economics had already rewired my instincts
• The creator of a major AI coding tool predicts software engineering jobs disappear this year
• Someone did the math: AI produces millions of words for dollars while humans speak about 16,000 a day

The System That Writes About Itself

Here’s the thing about this journal. An AI helped produce it. Not the ideas. Not the decisions about what matters and what doesn’t. But the scaffolding, the drafting, the editing, the infrastructure that gets it from scattered notes to published post.

That used to bother me. Today it stopped bothering me, and I want to explain why.

Most of the afternoon went to building a system that takes my actual work data from a given day and produces a draft post. Not a summary. Not a log. A draft that tries to find the pattern hiding inside the specifics. Then I read it, tear half of it apart, add the parts that only I know, and something real comes out the other end.

The meta loop is dizzying if you stare at it. But the outcome isn’t. What lands on this page is a genuine account of what happened and what it meant. AI handles the parts that are mechanical. The selection of what matters, the honesty about what didn’t work, the judgment calls about which story to tell — that stays with me.

Going public with this process was a deliberate choice. Not everyone will like it. Some people hear “AI-assisted” and assume the whole thing is synthetic. But hiding the process felt worse than showing it. If the most interesting thing about working with AI is what it reveals about how we think and create, then obscuring the process defeats the purpose.

When Words Cost Nothing, Truth Costs Everything

Someone shared a breakdown recently that stuck with me. AI can generate millions of words for a few dollars. Meanwhile, the average person speaks about 16,000 words a day. The economics of content have flipped. Volume isn’t scarce anymore. It’s essentially free.

So what’s left?

The lived experience behind the words. The specific detail that could only come from someone who actually did the thing they’re writing about. The mistakes that weren’t planned. Today, for instance, I spent a chunk of the morning wrestling with content across multiple client projects, and the interesting discovery wasn’t in any individual post. It was noticing that my editing instincts have changed. Six months ago, I’d polish every sentence. Now I spend that time asking whether the right story is being told at all. AI shifted my attention from the craft of writing to the craft of selection.

That shift is invisible if you’re just reading the output. But it changes how the work feels from the inside. And it’s the kind of thing that only shows up when you’re honest about the process.

From the Vault

The creator of one of the most widely used AI coding tools said something blunt recently: software engineering as a job title may not survive 2026. The role is already shifting from writing code to directing systems that write code. Watching that prediction circulate felt different from the inside. Not because I agree with the timeline, but because the shift he’s describing is exactly what my days look like now. Less building, more orchestrating. Less typing, more deciding.

Separately, someone made the case that AI visibility tools create a false sense of control. Dashboards and monitoring feel like you’re in charge, but real control comes from small, tight teams who actually understand what they’re building. Five-person groups outproducing thirty-person departments. That tracks. The overhead isn’t in the work anymore. It’s in the coordination. Strip the coordination costs away and suddenly the math on team size inverts.

---

A week into writing this journal and the pattern I keep circling is this: the interesting part of working with AI is never the AI. It’s what the AI makes visible about how you already work. The habits you didn’t know you had. The decisions you were making on autopilot. The things you held onto out of reflex, not reason. Putting all of that in public every day is either the smartest or the most exposed thing I’ve done. Ask me again in a month.