It's Getting Obvious Who's Actually Thinking

A pattern keeps surfacing in conversations with senior executives. The work is not bad. There is too much of it, it certifies nothing, and the people producing it are getting quietly weaker.

The scroll bar gives it away before you read a word. You open the document and the handle on the right is already a sliver, which tells you there are thirty screens stacked under the one in front of you, which tells you someone answered a question that wanted a sentence with a novella.

Lately I keep hearing the same thing from senior people, in different rooms, none of them comparing notes. Their staff are producing reports of a length and finish that would once have taken days, and the reports arrive in hours. Reading them, they feel something closer to fatigue than admiration. Not "this is impressive." Closer to "I am not going to read this, and I resent being asked to."

The interesting part is the resentment. These are not people who dislike effort. They have spent their careers reading effort off a page.

Here is what the page used to do. For the whole history of writing, a long, organized, fluent document was expensive. It cost time, attention, and a command of the subject that could not be faked cheaply. So we learned to treat the artifact as a receipt. The length was a receipt for hours. The structure stood in for thought. And the fluency told you a mind had been over this ground before. We were never reading the words. We were reading the price tag, and calling it intelligence.

Fluency was only ever a proxy. It was expensive enough to pass for the real thing.

The press is in every pocket now. The cost has gone to zero and the receipt has not changed its appearance: same length, same finish, same authority, none of the payment underneath. The document still looks like a receipt. It is a costume.

That is what they are reacting to, even if none of them puts it this way. They are being handed forged receipts and quietly appointed as the audit department. The aversion is not to the work. It is to being made to do the work the document was supposed to have done already: find the point, weigh the claims, pull the one useful sentence out of the forty packed around it. The labor did not vanish. It moved downstream, onto the reader, and multiplied on the way.

Think about what happens to a currency when anyone can print it. The notes do not get more valuable because there are more of them. They get less. You end up needing a wheelbarrow of them to buy a loaf of bread. Words are doing this now. Supply has gone vertical, the value per word has collapsed, and we are all standing in the bakery holding our wheelbarrows.

And the fix they reach for is older than email. Don't send it. Call me, or come and find me. Tell me the point to my face. Now look at who is asking. These are people with no time, who ration their attention in fifteen-minute blocks, and the thing they want is the most expensive thing in the building: a live conversation, in real time. Not a shorter document. No document. When someone that starved for time would rather you walk into their office than send them a file, the format is not the complaint. The document is.

It runs against the entire logic of the tools, which is worth sitting with. We built machines to spare us the friction of speaking to each other, to fire documents back and forth instead, and the first real result is that the documents became unreadable and people want to speak to each other again. The live voice turns out to be the one channel we have not learned to counterfeit at the speed it runs. You can generate a flawless memo. You cannot yet generate a flawless conversation in real time, with someone watching your face and asking the second question. Presence became the lie detector by default. Nobody designed it that way. It is the last surface that still costs something to fake.

Which leads to the harder observation.

More than one of these people has described the same reversal. They had decided, at some point, that a particular person was unusually sharp, on the strength of what that person wrote. Then they sat across a table from them, live, no draft to stand behind, and the sharpness was not there. The writing had been a front. Behind it was a model, and behind the model was less than expected.

Describe the mechanism, not the person, because the person is not the point and there are millions of them. Route your thinking through a machine often enough and the thinking stops happening on your side of the machine. The output still looks like cognition. The cognition has been relocated. A muscle grows only under the load you refuse to hand off, and this is a muscle people hand off all day, fluently, feeling more capable as it wastes.

There is an old anthropology for this. After the war, on some Pacific islands, people who had watched cargo planes land built runways out of straw and control towers out of bamboo and waited for the planes to come back. Every visible feature of the airfield was reproduced. The one invisible thing, the entire industrial apparatus that made planes land, was not, because it could not be seen and so could not be copied. A document generated this way is the straw runway. Every feature of intelligence is present except the part that produced it. The plane lands in the meeting, or it does not land at all.

None of this is an argument against the tools, and pretending it is would be its own kind of lie. They are the most capable instruments most of us have ever touched. The problem is narrower and stranger than "AI bad." It is about what we are handing over, and what that thing was doing for us while we still did it ourselves.

Follow the handing-over far enough and it stops being a story about competence and becomes one about strength.

Jordan Peterson puts a hard version of this, and the idea holds up apart from the noise that usually surrounds him. A harmless man is not a good man, he is a weak one. Virtue is not the absence of capacity. It is capacity held under restraint: the sword you know how to use and choose to keep sheathed. By that measure, when you let the machine do the difficult thing every time, you do not become restrained. You become incapable, and file it under efficiency. You have not sheathed the sword. You never forged one.

What the tools quietly remove is the friction that was the only road to getting good at anything: the hours of doing a thing badly, on your own, until you can do it well.

Incapacity is not a neutral resting state. It has a direction. Someone who cannot do hard things cannot carry weight, and life delivers weight regardless of whether you trained for it. The people who study where human cruelty comes from keep arriving at the same unglamorous source. Not strength gone bad. Weakness that could not bear what it was handed, soured into resentment, and went looking for someone to blame. The obvious answer to difficulty is to remove the difficulty, and it produces someone who cannot meet difficulty at all. You act out the same instinct when you swaddle a baby, and for a baby it is correct.

Now pull back from the individuals, because this was never a story about a few staff and a few executives.

A model produces the most probable next sentence. The most probable sentence is, by definition, the average one: the sentence closest to everything already written, the least surprising thing that could be said. Run that at a volume no human era could approach and you are adding to the shared record, layer on layer, the most forgettable possible version of every thought. Beige sediment, laid down faster than anyone can read it, each stratum indistinguishable from the one beneath.

This next part is speculation, so take it that way. A species that thinks through one instrument starts to lose the dialects of its own mind. Everyone's prose drifts toward the same competent center. A monoculture is the most efficient field in the world, identical rows, maximum yield, nothing wasted, right up to the morning the blight arrives and finds no variation anywhere to slow it. We are growing a cognitive monoculture and filing it under productivity.

And the exhaustion people report, the sense of falling behind a flood that never crests, is not a personal failing. It is an accurate reading of the situation. You cannot keep up. You were never going to. The flood has no cost function for your attention and no reason to stop. The volume is not evidence that you are slow. It is evidence that the thing producing it does not register whether you are there.

The reflex to reach for these tools is in everyone, the writer of this included. So the point is not a clean line between the people who use them and the people who do not.

The document was never the asset. The thinking was. We spent centuries unable to separate the two because the proof of work came bundled into every page, and now the proof of work is free and the two have come apart in our hands. We are about to find out, at scale and in public, who was doing the thinking and who was only holding the receipt.

But catching the people holding receipts is the small version of the question. The larger one turns on everyone. Every time you produce something now, you are answering it whether you mean to or not: does this add to anyone, or is it one more layer of chaff on a pile already too high to see over. Edify or erode. There is no neutral pile.

Where this ends is not clear, and anyone certain about it is selling something. The one trend that looks solid is the strangest. Output climbs while the capacity that used to sit behind it quietly thins, and the trouble with that trade is that it never feels like a loss. It feels like getting more done.

The meeting will tell us. It always did. That is why everyone suddenly wants one.