For years I measured my days by how full they felt. A calendar with no gaps. A Slack that never went quiet. A commit history that scrolled. If I ended the day tired, I told myself I’d done a good job. It took me an embarrassingly long time to notice that busy and useful are not the same thing — and that most of the work that mattered happened in the few quiet hours I kept trying to fill.

I don’t think I was lazy. I think I was hiding. The busy hours were the ones I could point to. The useful hours — the ones where I sat with a hard problem and barely moved — looked, from the outside and even from the inside, like nothing at all.

The days that felt full and produced nothing

I built a whole identity out of motion. I answered every message inside a minute. I volunteered for the meeting. I kept a dozen tickets in flight so there was always something to touch, always a small win to close before lunch. The blank editor made me anxious, so I filled it — with a refactor nobody asked for, with a config tweak, with anything that let my hands keep going.

And I read all of it as diligence. If someone had asked whether I was working hard, I could have produced the evidence: the closed tickets, the late-night messages, the plain exhaustion. What I couldn’t have produced was a straight answer to a different question — what actually changed because I was here today? Most days the honest answer was “not much, but I was very busy not-much-ing.”

The work that moved things was almost never the busy work. It was the quiet hour where I finally understood why the bug wasn’t where everyone was looking. It was the afternoon I spent not typing, deciding we should delete the feature instead of fixing it. Those hours didn’t feel productive. They felt like stalling. They produced the only things that mattered.

It almost never felt like work while it was working.

We stay busy to avoid the quiet room

Blaise Pascal wrote a line that has followed me around for years: “All of humanity’s problems stem from man’s inability to sit quietly in a room alone.” (That’s the popular compression; his actual passage in the Pensées is longer and gentler, but it lands the same.) His word for what we do instead was divertissement — diversion. We can’t stand the quiet, so we fill it. We go looking for noise and company and tasks and a game to play, precisely so we don’t have to sit with whatever the quiet would make us face.

I used to read that as being about leisure and mortality. Now I think it’s also about my job. The quiet room, for an engineer, is the moment when there’s nothing left to type and the problem is still unsolved. It is genuinely uncomfortable. You don’t know if your idea will work. You can’t tell whether the hour was well spent until much later, if ever. So I did the thing Pascal says we all do: I found something to keep my hands busy. Busyness was my divertissement. It let me feel productive without having to sit in the uncertainty where the actual thinking gets done.

Naming it that way changed how I saw a full day. A full day wasn’t neutral evidence of effort. Sometimes it was evidence of avoidance — a sophisticated, defensible, exhausting way of not doing the hard part.

Why busy looks like value, and useful doesn’t

The trouble is that the mistake isn’t only mine. It’s rewarded.

Two researchers at Harvard, Ryan Buell and Michael Norton, described something they called the labor illusion: we value work more when we can see the effort, even when that visible effort adds nothing to the result. Their case study was websites. A travel site that pauses to announce it’s “searching 100 airlines” gets rated as more valuable than one returning the same results instantly. Watching the labor made people trust the outcome more. The effort was theater, and the theater worked.

Engineering is full of that theater, and I performed it well. The visible grind — the fast replies, the packed meeting, the frantic afternoon — is legible to everyone around me. The useful part is invisible. Nobody can watch you decide not to build something. There is no commit for the hour you spent thinking your way out of a bad design. So the thing that’s easy to see collects the credit, and the thing that mattered collects none.

It runs deeper than perception. Silvia Bellezza, Neeru Paharia, and Anat Keinan found that in American culture, busyness has itself become a status symbol — we now read a person’s lack of free time as a signal that they’re competent and in demand, the way an earlier era read a life of visible leisure. “Slammed” is a humblebrag. I said it about myself constantly, and I meant it as a boast. I was advertising the exact thing I should have been suspicious of.

AI ate the busywork, and I felt strangely idle

Here is what finally forced me to confront this instead of just nodding along with Pascal: the busywork left.

For most of my career, busy and useful were tangled together, and I think that tangle is why I confused them for so long. A lot of genuinely useful work required grind. You couldn’t reach the interesting decision without first typing out the boilerplate, reading the docs, wiring up the fifth near-identical endpoint by hand. The grind was a toll you paid to get to the thinking — but it also filled the day, and it felt like the work, because physically it was most of the work.

Now I code with AI tools every day, and the toll is mostly gone. The boilerplate writes itself. The unfamiliar API gets explained in place. The mechanical eighty percent that used to eat my afternoons collapses into minutes. And the first time I had a genuinely productive day this way — more shipped, more actually decided — I noticed I felt oddly idle. Guilty, even. My hands hadn’t moved enough. The tiredness I used to cash in as proof of a good day never showed up.

That feeling was information. It told me how much of my sense of being a good engineer had been resting on motion — on the grind, not on the judgment the grind was supposed to serve. When the machine took the busywork, it didn’t just make me faster. It removed the thing I’d been hiding behind and left me standing in Pascal’s quiet room with the part of the job I’ve always found hardest: deciding what’s worth building, noticing when correct-looking code is subtly wrong, holding the whole shape of a problem in my head. The tools are extraordinary at the busy. They are not going to sit in the uncertainty for me.

This is the mirror image of something I’ve written about before — what you quietly lose when AI writes your code is the learning that lived inside the struggle. What you see, once the struggle is gone, is how small the useful part of the day always was. Both are the same fact viewed from opposite sides: the grind was never the point, and I’d spent a long time treating it as the point.

I won’t pretend to know exactly where this goes. But the direction seems clear enough to say out loud, carefully. As the grind keeps shrinking, looking busy stops being available as a substitute for being useful. There won’t be much visible labor left to perform. What remains is the quiet, unmeasurable judgment — which I’ve come to think is most of what seniority is — and that was the point all along. I couldn’t see it under the pile of work that used to bury it.

When busy is the actual job

I want to be careful here, because “busy is avoidance” can curdle into its own kind of pose — the engineer too enlightened to answer their messages.

Sometimes busy is exactly the work. An incident at 3 a.m. is pure motion, and every bit of it is useful. A real deadline sometimes demands a stretch of head-down volume, and that’s not avoidance, that’s delivery. When you’re genuinely new to something — a language, a domain, an unfamiliar codebase — you need the reps, and there is no shortcut through the hours; the grind is how the understanding gets built in the first place. The operational work that quietly keeps systems alive is unglamorous, invisible, and completely essential.

So the point was never that motion is bad, or that thinking is nobler than doing. The point is narrower and harder. I stopped letting the feeling of being busy stand in for the evidence that I’d been useful. Busy is honest when I can point to what it produced or why it was necessary. It’s avoidance when its main product is the feeling of having worked. The two are identical from the inside — same fatigue, same full calendar, same clean conscience. The only way I’ve found to tell them apart is to ask, when the day is over, the question I used to dodge: not “was I busy,” but “what is different now because I was here.”

What counts as a full day now

I still fail this most days. When the quiet work scares me — when I genuinely don’t know whether the idea is any good — I still feel the old pull to open the calendar and fill it, to go find something safe to touch. The difference now is that I catch myself, and I know what the reaching is. Not diligence. Diversion.

So I’m trying to build a different habit. Measure the day by what changed, not by how tired I am. Treat one quiet, thinking hour as the most productive thing I did, even though it will never look like it and no one will ever see it. Let the machine take the busywork, and spend the space it opens on the part only I can do — instead of rushing to refill that space with more motion, which is, I’m realizing, the oldest reflex I have.

I don’t have this figured out. I’m just no longer willing to call myself useful on the grounds that I was busy.

Sources