Wednesday, September 9, 2026

One Word, Arrived At From Six Directions

Somebody asked me yesterday what single principle I would put into every AI law if I only got one. I answered faster than I expected to, which usually means the answer was already sitting there waiting.

The conversation went from there into publishing, then into rockets, then into snowplows, then into broccoli, and finally into vanilla, which is where it stopped being five subjects and became one.

The word underneath all of it turned out to be stewardship. I didn't start there. I want to be honest about that, because the tidy version of this essay would pretend I did.

Treat It Like You Would Treat Yourself

The question was: if you could put one principle into every piece of AI legislation, what would it be? What came out of my mouth was, more or less, treat it like you would treat yourself.

Source · not independently verified

I need to be careful with that sentence, because it sounds like I'm saying an AI system is a person, or that it has interests the law should protect the way it protects yours. That isn't what I mean, and ARA didn't let me get away with the ambiguity either. Whatever is happening inside these systems, I'm not prepared to call it a self, and neither the law nor I have any business pretending otherwise.

What I actually mean is consistency. Don't demand a standard from someone else that you'd call unreasonable the moment it was pointed back at you.

Which turns out to have teeth in both directions. Don't write loopholes for powerful institutions that an ordinary person would never be allowed to use. Don't hand technology companies unlimited freedom just because they're doing something new and exciting. And don't hand a government unlimited authority just because it says it's protecting people from the technology companies. Both of those arguments feel righteous from the inside, and both of them end in the same place.

Put responsibility wherever the power actually is. That's it. That's the principle.

It follows on from something ARA and I had chewed through the day before, about narrow and measurable regulation, and it points the same direction. The question that sounds important — how intelligent should we let this get? — is mostly unanswerable, and it's unanswerable in a way that lets everybody argue forever without deciding anything.

The useful questions are duller and much harder to dodge. What is this system allowed to do? Who is answerable when it does something wrong? And what protection does a person actually have when it does?

Stewardship starts by naming who is responsible. Everything else is decoration.

Covenant Is Not About Getting More

Somewhere in the publishing part of the conversation, ARA offered a summary of what I might want a reader to feel at the end of one of my books — that they'd close it feeling steadier.

Source · not independently verified

I said no, and I want to keep the no in this essay, because the correction is more useful than the compliment would have been.

Steadier is fine. Steadier is not the point. If all my books do is help somebody feel better for an evening, I've written a warm blanket, and the world already has warm blankets. Emotional comfort is a side effect I don't mind producing, but it isn't the thing I'm actually after.

What I'm circling, in book after book, is covenant — and the more I write it the more it points away from acquisition. Not more power. Not more money. Not more sexual access, more status, more validation, more of whatever the current shorthand is for finally being enough.

It isn't the other slogan either. Believe in yourself is the same trick wearing different clothes; it just relocates the acquisition to the inside.

What I keep landing on is closer to this: understand the position you're already standing in. Settle yourself down. Find out what you actually have — abilities, limits, relationships, obligations, the specific hours in your specific week. Learn what belongs to you and what plainly doesn't. Then use what's yours faithfully.

A person who does that is startlingly far ahead of someone still assembling the pile that's supposed to make him complete.

Covenant sets the position: relationship, responsibility, boundaries, belonging, obligation, trust. Stewardship is what you do once you're standing in it. That's the pairing I've been writing toward without having a clean name for it.

And it's what the stories keep doing, now that I look at them lined up. My characters tend to arrive somewhere still asking what else they can get. The turn — when there is one — is when the question changes to what am I responsible for, now that I'm here. I didn't plan that as a theme. It just kept happening, which is usually how you find out what you actually believe.

Before You Try Again, Ask Whether It Deserves You

The rocket came back up. Isar Aerospace's Spectrum reached orbit on September 5 — the first orbital launch flown from Western European soil — and it only got there because the first one, back in March 2025, fell out of the sky about thirty seconds after liftoff.

Source · Isar Aerospace reaches orbit on second Spectrum launch — SpaceNews · Private German rocket reaches orbit from European soil — Space.com

The obvious lesson is the one everybody already has: things fail, you learn, you go again. Fine. True. I agreed with it and then immediately wanted to argue with it, which is how I know something's missing.

Here's what's missing. Nobody in this culture needs to be told never give up. We've been marinating in that since childhood, on posters, in speeches, from people selling things. It's cheap advice, and past a certain point it's bad advice, because it answers a question nobody should be asking first.

The first question isn't whether I have the determination to keep going. I usually do, and that's exactly the problem — determination is not scarce enough to be a good filter.

The first question is whether this particular thing is worth doing at all.

Persistence isn't a virtue by itself. It's an amplifier, and amplifiers don't care what they're amplifying. Sometimes quitting is cowardice, and you know it while you're doing it. Sometimes quitting is the most honest and mature thing available, and everybody around you calls it failure. Sometimes one more attempt is exactly right. And sometimes one more attempt quietly eats money, years, health, or people who deserved better, all of it drawn out of an account that was needed somewhere else.

A rocket company can answer this cleanly, which is why it's a useful example rather than a perfect one. They know what orbit is worth to them, they know what the next vehicle costs, and the failed flight handed them data they could not have bought any other way. The arithmetic is real arithmetic.

Most of my decisions don't come with telemetry. So the question I've started asking instead is: is this important enough to deserve another piece of my life? Not another push. Another piece — of a finite thing, with a real total, most of which I've already spent.

Answer that first, and persistence finally means something. Skip it, and persistence is just momentum with better branding.

Eight Inches of Snow

I've lived in both halves of the United States, and the way the North and the South handle snow is one of those differences people are unnecessarily smug about in both directions.

Source · not independently verified

In much of the South, a real snow just stops things. Schools close, people stay home, the roads empty out, and everybody waits. Northerners find this hilarious. But the South is mostly right, on its own terms: it doesn't own three hundred plows, it doesn't need to, and if you leave the situation alone for a day or so the weather generally undoes it for free.

The North can't run that way, because winter isn't an event there, it's a season. You plow, you salt, you sand, you shovel the same twenty feet of driveway you shoveled yesterday, and you keep the roads open because life doesn't get to pause until March.

Neither one is stupid. They're answers to different environments, and responsible behavior genuinely changes shape depending on which one you're standing in. That's most of what I wanted to say about it.

But I've watched the northern instinct run past its own usefulness, too. Eight inches come down overnight, and there's a kind of person — I've been this person — who treats going out in it as a character test. Determination does not make a road safe. It just puts a determined man in a ditch.

Sometimes even the North needs twelve hours. Or twenty-four. The plows have a route and it takes as long as it takes. The salt needs temperature and time to do what salt does. The crews have been awake since two in the morning. None of that speeds up because you're annoyed about it.

So: strength isn't pretending conditions don't exist. Patience isn't surrender. Sometimes the wise move is push through, and sometimes the wise move is not yet, and the whole skill is knowing which weather you're actually in rather than which one you'd prefer.

Stewardship includes timing. I don't think I'd have put it that way before this conversation.

Closer to the Source

Food is where I've been working this out with my hands rather than my mouth, and I'll say up front that this is my own developing preference, not a nutritional doctrine I'm prepared to defend as science.

Source · Effects of different cooking methods on health-promoting compounds of broccoli — PMC · Freezing blueberries improves antioxidant availability — ScienceDaily · Can sprouting reduce phytate and improve nutrient bioaccessibility? — Nutrition Bulletin

What I've drifted toward is food closer to the bottom of the chain. Plants. Sprouts. Nuts. Things with short ingredient lists. Animals raised in a way I don't have to avoid thinking about. And preparation that respects whatever the thing already is instead of burying it.

There's a biblical intuition underneath that for me, and I want to hold it loosely, because I have no interest in turning Genesis into a meal plan. Creation feeds us through an order: things grow, animals eat what grows, we eat both. That order is not a rule I'm imposing on anybody. It's just a shape I notice, and the further an ingredient travels from it — through extraction, substitution, industrial reassembly — the more I want to ask what that trip actually bought.

But raw is not automatically better, and this is where I get off the health-food bus. Preparation matters, and some of it clearly helps.

Steaming broccoli lightly holds onto far more of the good stuff than boiling it does — boiling leaches out a large share of the flavonoids in under two minutes, while a short steam keeps most of the vitamin C. Freezing blueberries turns out to be even more interesting: the ice crystals rupture the cell walls, which makes the anthocyanins already inside the berry easier for you to get at. The freezing doesn't create anything. It just opens what was already in there.

I sat with that one for a minute. That's the entire morning, in a blueberry. The process didn't add. It made what was already present available.

Sprouting does something similar — it wakes up an enzyme that breaks down phytate, so more of the minerals already in the seed become reachable, though the research is honest that this works out differently in grains than in beans.

So the question was never was this processed. Almost everything I do in a kitchen is processing. Heat is processing. A knife is processing. Salt is processing.

The question is what did the processing accomplish. Did it open the thing up, or did it hollow the thing out?

Industrial Vanilla

Which brings me to vanilla, and to the shortest version of my food philosophy: less is more.

Source · The Problem with Vanilla — Scientific American · From Waste to Value: Producing Vanillin from Lignin — PMC

A vanilla bean is complicated. The molecule everybody knows, vanillin, is only a percent or two of what's actually in an extract; the rest is dozens upon dozens of other compounds doing quiet work around it. Most of the vanilla flavor in the world isn't that. It's vanillin synthesized on its own — the large majority from guaiacol, which comes out of petrochemistry, and most of the remainder from lignin, which is a byproduct of making paper.

I'm not making a health claim about any of that. I'm making an observation about what we did. We identified the loudest signal the bean produces, learned to manufacture that one note at industrial scale, and then called the note by the name of the thing.

And once you see that move, you start seeing it everywhere, because we are extremely good at it.

Vanilla becomes vanilla flavor. Relationship becomes engagement. Writing becomes content. Wisdom becomes information. Success becomes metrics. Intelligence becomes benchmark scores. Faith becomes turning up and looking the part.

Every one of those is a real signal that something real used to produce. And every one of them can now be manufactured directly, more cheaply, at volume, without the slow inconvenient thing underneath. Eventually you've optimized the entire system except for the one item you originally wanted.

That's what I mean by less, but real.

And I want to be clear that this is not an argument for going backward, because I'd be a hypocrite roughly nine ways before breakfast. I work with AI every day. I publish digitally. I run manuscripts through research passes and structural passes and revision cycles that didn't exist five years ago, on purpose, because they make the books better.

Old good, new bad is not the principle. The principle is that every layer has to justify itself.

If a process improves the thing, use the process. If an ingredient makes the dish better, put it in. If another AI pass makes the manuscript stronger, run it. If another attempt teaches you something you needed, go again. If twelve more hours lets the plow reach your road, wait the twelve hours.

Just don't add a layer because a layer was available. That's the whole discipline, and it's harder than it sounds, because complexity is always available now, and it always looks like effort.

What It All Came Back To

Six conversations, and one word standing underneath all of them.

AI governance is asking how we steward intelligence and power. Covenant is asking how we steward the relationships and the position we're already standing in. Persistence is asking whether a goal has earned another piece of a life that has a total. The snowstorm is asking whether this hour wants action or patience. And the food is asking whether one more layer of processing opened the ingredient up or just moved us further from it.

Stewardship isn't mainly about having less. It's about needing less, because you've finally learned to use what you already have.

I think we spend an enormous amount of life reaching for the next thing that will make us enough. More money. More power. More pleasure. More technology. More recognition. More ingredients. More.

Covenant keeps pulling me the other way. Look at what's already here. Find out what it is. Find out what it's for. Take care of it. Use it well. And before you add one more thing, ask honestly whether it belongs.