Saturday, September 12th, 2026

First real week back. The kind where you get to Friday and can't quite account for it, not because it was bad but because it was dense, and dense weeks have a way of using up the part of your brain that notices things.

So I'm sitting down Saturday morning to notice a few things on purpose. Coffee's hot. Sheldon has claimed the exact patch of floor where the sun comes through, which he does with the confidence of an animal who has never once questioned whether he deserves it.

Here's what's been rattling around.

This came up sideways.

Somebody asked me a question this week about something I've done a hundred times, and I answered it fast, in that flat certain tone that means the answer arrived before the thinking did. And then twenty minutes later, on my own, I went back and asked myself how I actually knew that.

The honest answer was that I didn't know it. I'd just done it enough times that being unsure had stopped being available to me.

That's not knowledge. That's a groove.

Once I noticed it I couldn't stop finding it. I went through the things I'm most certain about in the business and started sorting them into two piles. Pile one is things I'm sure of because I've examined them, tested them, been wrong about them and corrected. Pile two is things I'm sure of because I've been doing them since a period when they worked, and I've never gone back to check whether the conditions that made them work still exist.

Pile two is much bigger than pile one. It's not close.

What makes this uncomfortable rather than just interesting is that the two piles feel identical from the inside. Real conviction and worn grooves produce the exact same sensation. Same flat certainty. Same slight impatience when somebody questions it. There's no internal signal that distinguishes them, which means you can't feel your way to the difference. You have to go looking, deliberately, in a mood where you're willing to find something you don't like.

I've been trying an exercise on this and I'll pass it along, though I want to be clear it's about a week old so I'm not claiming much for it. When I catch myself answering something instantly, I write the answer down and then I write the last time I actually tested it. Not the last time I used it. The last time I checked whether it was still true.

Most of the dates are old. A few of them, I can't produce a date at all, which is its own kind of answer.

The version of this I'm still chewing on is the one I don't have a good handle on. Experience is supposed to be the thing that makes you better. And a lot of the time it is. But the mechanism by which experience makes you better is also the mechanism by which it makes you stop looking, and I don't know how you keep one without getting the other. Deliberately doubting everything is not a workable operating system. You'd never get anything done and you'd be exhausting to be around.

So maybe the move isn't doubting more. Maybe it's picking three or four things a year, the load bearing ones, and putting them on trial on a schedule whether or not anything's gone wrong. Not because they're broken. Because you can't tell from the inside whether they are.

I don't have that system yet. I've got the observation. That's usually where these start.

Different thread, and this one's less comfortable.

I was going through my phone this week looking for a number and ended up scrolling further than I meant to, the way you do. And I kept hitting names of people I would describe, without hesitation, as friends. People who mattered. People who were in the room for things.

Then I'd look at the last message and it would say something like February. Or last year.

Not one of those relationships ended. There was no falling out. Nobody did anything. They just went quiet, in the specific way things go quiet when everybody involved is busy and nobody's the one whose turn it obviously is.

Here's the part I've been sitting with. In the business, I'm relentless about maintenance. I know that systems degrade without attention. I've built entire frameworks around the idea that you can't set something up and walk away and expect it to still be running in a year. I say some version of this to people constantly.

And then I've been running my actual relationships on a model I would laugh at if a client proposed it, which is: it'll be fine, we go way back.

We go way back is not a maintenance plan. It's a description of the past.

What I think happened, and I'm being honest here rather than flattering, is that I built a hierarchy in my head without ever deciding to. At the top, things with deadlines. Then things with money attached. Then things that would be noticed if they were missing. And relationships with old friends sit near the bottom of that stack, not because they matter least, but because they're the only category that doesn't complain. Nobody sends you a reminder that you haven't talked in eight months. There's no overdue notice.

So it's the one thing in my life with no forcing function, which means it's the one thing that runs entirely on whether I happen to think of it, which means it runs on nothing.

And then the longer it goes, the higher the activation energy gets. Reaching out after two weeks is nothing. After eight months it feels like it requires a reason, an occasion, an apology. So you wait for one. And no occasion comes, because occasions are things that happen between people who are in touch.

I'm not going to turn this into a system. Something in me resists putting my friends on a recurring task, and I think that resistance is worth respecting even if it's slightly irrational.

But I've been thinking about the difference between systematizing a relationship and systematizing the reminder. Those aren't the same thing. The conversation doesn't have to be scheduled. The moment where I glance at the list and go, oh, right, him, that one can be.

Started this week with three people. Two wrote back within a day, easy, like nothing had happened, which is both wonderful and slightly damning, because it means the barrier was entirely in my head and had been for months.

Third one hasn't written back yet. That's fine. That's the risk you accept when you let something sit.

This one's been building for a while and I think it finally has a shape.

I make a lot of things in public. Four letters a week, a podcast, posts across enough platforms that I've had to build machinery to keep track. And the machinery is honest work and I'm not apologizing for it. But there's a question underneath it I've been avoiding, and it surfaced this week when I finished something I was pleased with and immediately, before I'd even sat back, started thinking about where it would go and how it would land.

The gap between finishing and posting was maybe four seconds.

Which made me wonder what I'd have done with it if there were nowhere to put it.

So I've been running a thought experiment. If tomorrow everything I make became invisible, no audience, no reach, no possibility of anyone ever seeing it, what would I still make?

The answers came back cleaner than I expected, which was the surprise.

I'd still write. Not the same way and not as often, but I'd still write, because writing is how I find out what I think and that function doesn't require a reader. I know this because I've got years of things nobody's ever read and I wrote them when there was no version of my life where anyone would.

I'd still build systems. Absolutely and without hesitation. The pleasure of a thing that runs correctly is completely private. Nobody has ever needed to see one of my automations for me to enjoy that it works.

I'd still help specific people with specific problems. That one held up under scrutiny. The satisfaction of watching somebody's actual situation improve doesn't route through an audience at all.

And then there's a category that didn't survive the test, and I'm not going to pretend otherwise. Some of what I make exists because it will be seen. Not most of it. But some. And the useful thing isn't to feel bad about that, it's to know which is which, because they need completely different treatment.

The stuff I'd do anyway is the foundation. It doesn't need motivation, it doesn't need a strategy, it just needs to not get crowded out. The stuff that exists to be seen is a business function, and business functions should be evaluated like business functions. Is it working. Is it worth the hours. If not, kill it, and kill it without any of the emotional weight you'd feel about killing the first kind.

The failure mode, and I've lived in it, is treating them as one thing. When you can't tell which is which, a quiet week makes you question the whole enterprise, including the parts that were never about attention in the first place. A post that doesn't land shouldn't be able to make you doubt whether you should be writing. But it can, if you never sorted them.

Here's the thing I keep circling back to. I think the test isn't really about audience at all. I think it's a way of asking whether the thing has an inside.

Some work has an inside. There's something happening in it for you, independent of the response, some pleasure or curiosity or usefulness that would survive being ignored. And some work is hollow, engineered entirely for a reaction, and it can be perfectly good work, technically excellent, well made. It just won't hold you up when the reaction doesn't come.

You need some of both. I'm not arguing for purity here, that's a luxury position and I'm not in it. But if the ratio tips too far toward hollow, you end up in a business that only feels worth doing when it's going well, which is exactly backwards, because the whole point of building something durable is having something to stand on when it isn't.

That's the week.

Familiarity that's been passing itself off as certainty. A maintenance schedule I apply to everything except the people. And a question about what would survive if the audience vanished, which turned out to have a better answer than I was braced for.

None of these came with solutions attached. That's not what Saturdays are for. Tuesday's for solutions.

Today's just for looking at the thing honestly and letting it sit there.

Sheldon's moved with the sun. Third time this morning. He does not appear to be examining any of his assumptions and he seems fine.

One step, one day. Grace over guilt. — Dan Kaufman