Every Sentence You Generate Faster Than You Think Is a Bill Someone Else Has to Pay

Last week, I wrote a newsletter and decided, after reading it back, to put it in the drawer. The argument was there and the references checked out. But somewhere between writing the thing and reading it I had crossed a line I could not quite name, and what was in front of me was something I did not want to put into the inboxes of people whose attention I had asked for.

The advice I had absorbed over years of doing this told me to send it anyway: don’t let perfect be the enemy of good; the reader rewards regular publication more than they reward the occasional polished piece. I had believed all of it. Most of the time it had served me well.

But I still did not send it out, and the cadence broke. And in the week that followed I noticed that the advice I had taken as a principle was conditional, and the condition was no longer holding.

“Don’t let the perfect be the enemy of the good.”

Voltaire is usually credited, though the version we live by comes from a long succession of productivity writing that, with reason, argued that the main thing standing between most people and useful output was their own throat-clearing. Anne Lamott named the corrective in Bird by Bird: the willingness to write a shitty first draft. Whole shelves of advice were written to give people permission to put unfinished work into the world.

It was good advice, for a particular reason, and the reason is worth saying out loud, because the reason is what changed. The advice assumed that the limiting factor on good work was the writer’s own resistance to letting it leave the desk. Production was costly; judgment was cheap. So the corrective leaned on the side where the friction was.

That assumption no longer holds.

The economist Herbert Simon saw the shape of the problem in 1971, long before anyone had the means of producing prose at the speed we now can:

“What information consumes is rather obvious: it consumes the attention of its recipients. Hence a wealth of information creates a poverty of attention.”

Simon was writing about an organisation drowning in reports. He had no idea what was coming. The recent tools have not broken his law; they have poured the contents of every research library and every almost-finished thought into the same channel where a reader still has only the same twenty minutes.

When production becomes free, what is scarce on the other side of the equation is judgment. Oliver Burkeman, who has been writing about this with more patience than most, calls it the efficiency trap: every tool that promises to help us do more ends up producing more for us to do. The harder we work to get on top of the inbox, the more the inbox grows. The same dynamic now applies to writing. A tool that lets me produce a competent draft in twenty minutes does not free me from the question of whether the draft is worth producing.

Researchers at Stanford and BetterUp have a name for what fills the channel when nobody asks the question. They call it workslop: writing that looks adequate but has had the actual work of thinking offloaded onto whoever receives it. Their survey found 40 percent of workers had received it in the past month. Recipients lose hours sorting through the stuff and end up viewing the senders as less reliable than they had been before.

Byung-Chul Han has been making a broader version of the point for a decade. The pathology of the modern subject, he argues, is the inability to stop producing. The achievement society replaces external command with internal compulsion that never lets up. To choose not to produce something has become an act that takes more self-knowledge and self-management than choosing to produce it.

The draft I put back into the drawer was a small instance of this. I had typed it out with a model open in another window, and what had been generated had the shape of writing without the substance of thought. The argument was there. But it was stale. Every paragraph said something I had said before, dressed in a form that was a bit too smooth. I read it twice and could feel myself wanting to send it anyway because Tuesday was Tuesday and the weekly cadence was the weekly cadence. Then I read it a third time and noticed that what I would regret more than missing a week was the version of myself who would have pressed send on this. So I closed the file.

Consistency was a fix for a real problem: the writer’s resistance to letting unfinished work leave the desk (and I think you can replace ‘writer’ with ‘knowledge worker’). What we now have is a different problem — what to do when output is free and the only remaining bottleneck is judgment — and the old advice has nothing to say about it.

I wrote a piece a while ago called “The Responsibility to Focus.” Most of it was about protecting your own attention from people and platforms trying to take it. A shorter section at the end argued that if you have any audience at all, even a small one, you also carry some responsibility for theirs. You direct attention whether you mean to or not. That section was the right idea half-developed. This is the rest of it.

When the act of writing was expensive, restraint took care of itself. The cost of the act filtered most of what should not have gone out. Now the filter is gone. So the question of who should send what to whom, and when, and why, has stopped being answered by the friction of the medium. It has to be answered by whoever is doing the sending.

And the answer is moral, not just aesthetic. A reader’s attention is finite. If they give you fifteen minutes of it, those minutes were taken from somewhere that mattered: from the work they were doing, or from the work they have been meaning to do that might actually move their life forward; from the children they were trying to be more patient with, or the partner they had not really listened to that morning; from the kind of thinking that only happens when nobody is asking anything of them at all. To ask for those minutes and hand back something you knew was thinner than your usual work is a way of taking from them. The cost you saved by not thinking shows up on someone else’s bill.

The German philosopher Josef Pieper, writing in 1948, argued that what holds a culture together is not its rate of production but its capacity for what he called leisure, by which he meant something close to the willingness to stop. A society that cannot stop, he thought, ends up unable to do its serious work, because the serious work requires the silence in which one can hear oneself think. He was making a deeper point than the modern reader sometimes hears: it is not only the person who cannot stop who suffers from a culture of unceasing output. It is everyone who has to live in the noise that person produces.

The slop problem is not contained in newsletters. It is in the Slack thread where the contribution was generated rather than thought through. It is in the strategy document that reads coherently but commits to nothing. None of these on their own are catastrophes. Each is a small taking. A culture in which everyone takes a little, all the time, from everyone they touch, becomes one in which nobody can think clearly about anything important, because the inputs to thinking have been poisoned at the source.

The discipline that replaces consistency is harder to talk about because from the outside it looks like nothing. It is the discipline of not adding to the noise on the weeks when the only thing you could honestly add is more noise.

The old advice was: do not let the perfect be the enemy of the good. The new discipline asks something trickier. Before you press send on anything you generated faster than you thought, ask whether the person on the other end would have agreed to spend their attention on it had they known how little of yours it cost.

Yes, that question can kill work that looks ready to send.

But that is what it is for.