For roughly a thousand years, Chinese women had their feet broken and bound — toes folded under the sole, wrapped in tight cloth, reshaped over months into a tapered point about three inches long. The practice began in the Song dynasty and persisted well into the twentieth century. It was excruciatingly painful, permanently debilitating, and nearly universal. A woman with unbound feet was, in practical terms, unmarriageable.
If you were a young woman in, say, 1850 — smart, clear-eyed, fully aware that this tradition was cruel — what would you hear when you voiced that awareness?
“That’s just how things are. You can either bind your feet and have a life, or don’t and accept the consequences.”
This response is factually correct. It’s also a perfectly useless thing to say. It takes the current state of the world and presents it as a boundary condition — a fixed constraint to be optimized around, not a problem to be questioned. The person saying it isn’t wrong. They’re just done thinking too early.
I’ve started calling this pattern deal-with-it-ism: the reflexive treatment of a broken system as a fixed landscape — something to be navigated, never questioned. It’s the instinct to skip past “should this be the case?” and jump straight to “given that it is, what do you do?”
The instinct to adapt
We all do this. It’s not a character flaw — it’s a survival instinct. You don’t philosophize about the nature of fire while your kitchen is burning. Deal-with-it-ism is genuinely useful in a huge number of situations, and the people who are good at it tend to be the ones who keep teams running and projects shipping.
The problem is when it becomes the only mode of thinking available — when every observation that the system itself might be broken gets flattened into advice for how to operate within it.
Point out that a game’s difficulty spike is poorly designed? “Get good.” Raise a concern about architectural debt in a codebase? “We need to ship. File a ticket.” Question whether a hiring process selects for the wrong traits? “That’s the process. Prep better.”
The pattern is always the same: the system disappears from view, and what remains is you — and your apparent inability to deal with reality. A critique of the terrain gets repackaged as a deficiency of the hiker. The system you’re optimizing within is itself beyond question.
Coherence as progress
Here’s what I think is actually at stake.
Consider the arc of civilization, and ask: what actually changes? Technology changes, obviously. Laws change. Norms change. But underneath all of that, there’s a subtler, slower process: we get better at noticing when things don’t cohere.
Sometimes that’s a contradiction between values. We used to believe in human dignity and watch gladiators kill each other for entertainment. We used to believe in equality and practice slavery. We used to believe children deserve protection and send them into coal mines. Each of these was resolved by someone noticing: these two things we both believe cannot both be true.
Sometimes it’s a gap between capability and reality. We had the technology to put a computer in every pocket for years before anyone built the thing that made people actually want one. The pre-iPhone world wasn’t morally inconsistent — but it was incoherent. What we could do and what we were doing didn’t match. Someone felt that tension before they could articulate it, and the articulation became a product.
Sometimes it’s simpler still — just the persistent, nagging sense that something could be better than it is, without a clear framework for why. A designer plays a hundred games that handle inventory the same way, and something about it has always bothered her, and one day she builds something different and it’s obviously right.
The common thread across all of these isn’t the specific domain. It’s the sensitivity: the ability to feel that two things don’t cohere — values and actions, capability and reality, what-is and what-could-be — before you have the language to prove it.
This is also why every product that genuinely succeeds is, at some level, the resolution of an incoherence. Sometimes it’s a market gap identified through analysis, but more often than not, a felt gap, experienced by someone who lived inside the problem long enough to sense the shape of what was missing. The need existed before the product did. The product was the articulation.
This is what I’d call wisdom in the civilizational sense — not the accumulation of knowledge, but the progressive elimination of incoherence. Not only the moral kind — the “we say we believe X but we do Y” kind — but also the generative kind: the distance between how things are and how they could be. There’s a mathematical elegance to this: in formal logic, a system with a contradiction can prove anything — ex falso quodlibet. Every incoherence you eliminate is one fewer license to rationalize, one fewer way for a bad argument or a bad product to survive unchallenged. Progress, in this framing, is the slow tightening of what we can say with a straight face — and the slow expansion of what we can build with a clear conscience.
The formalizable and the not-yet-formal
That sensitivity — the one that resolves incoherence — is, by definition, not yet formalized. If it were, it wouldn’t be sensitivity; it would be a checklist. And this is exactly where the ground is shifting.
Over the past two years, a specific kind of professional confidence has been collapsing: “Sure, AI is impressive, but it can’t do X.” Write production-quality code. Understand a large codebase. Work autonomously with real tools. Each X falls, and the goalpost moves again. The pattern isn’t that AI is getting smarter in some general sense. It’s that everything formalizable is falling — clean code, system design, architectural patterns — not because it’s easy, but because the knowledge already exists in text. AI reads text. The gap closes.
What doesn’t fall is the sensitivity itself. And the reason, I think, is a hard limit — not a temporary one.
AI can write a song that most people will call good — because “good” has already been defined by a century of popular music. It can optimize within an existing taste space. But it can’t create a new taste space, because that would require validation from an audience that doesn’t yet know it wants something different. Jazz wasn’t validated by ragtime listeners. Punk wasn’t validated by prog-rock fans. The new genre has to find its people — and that finding is a human process, not a statistical one. A product isn’t valuable because it’s novel. It’s valuable because people used it and their lives got less incoherent.
The validation function is human. It always has been and it doesn’t scale.
The pipeline is: someone feels incoherence, struggles to articulate it, eventually formalizes it. AI can accelerate every stage except the first and the last. Generation is becoming free. But the bookends — the spark and the verdict — are always human.
The trap
Now look at what deal-with-it-ism actually does in this light. It’s a suppression mechanism — aimed precisely at the faculty that matters most. Someone feels an incoherence that hasn’t been formalized yet, and deal-with-it-ism says: stop. Don’t feel that. Adapt.
It takes the one thing that can’t be automated — the spark, the felt sense of “this doesn’t cohere” — and trains you to ignore it. Not through argument, but through social pressure. You learn that noticing brokenness is naïve. That questioning the system is a luxury. That the mature, serious, professional thing to do is to navigate the terrain as given and save your energy for things you can control.
This was always a bad trade. But it was at least a legible one. The system-navigators had real skills — hard-won, valuable, respected. They kept things running. They shipped. In a world where execution was expensive and generation was slow, knowing how to work within constraints wasn’t just practical; it was the whole game.
That world is ending.
Generation is becoming free. The formalizable is being formalized at a rate that makes “I know how to navigate this” a depreciating asset. The person who memorized the codebase, the one who knew every workaround in the pipeline, the one whose value was fluency in a broken system — that value had a shelf life, and it’s running out.
What doesn’t depreciate is the thing deal-with-it-ism was designed to suppress: the ability to feel that something is wrong before you can prove it. The willingness to stay with that discomfort instead of explaining it away. The stubbornness to keep thinking when everyone around you has stopped.
Deal-with-it-ism told you to stop caring about the things you couldn’t fix. That was always bad advice. Now it’s obsolete advice. The thing that will matter most — the thing that already matters most — is the part of you that refuses to look at something broken and call it fine.
Start caring.