On Doing Nothing
The discipline almost nobody schedules
I want to make a case for a working practice that almost no software team, founder, or product organisation takes seriously: blocks of time spent doing nothing in particular. Not meditation. Not deep work. Not journaling, walking with a podcast, or any of the lightly disguised productivity rituals that have absorbed the word in the last decade. Just sitting. Embracing boredom. Letting the part of you that screams “be productive, justify your existence” scream itself hoarse, and refusing to obey it.
There are two arguments for this. The first is instrumental. Doing nothing is one of the most productive things a knowledge worker can do, and the science is embarrassingly clear about why. The second is more important and harder to defend in a working culture. You have the right to exist as something other than a productive unit, and the contrary assumption is a piece of inherited industrial-age thinking that has no business shaping how a software career is lived. Both arguments are real. I will take them in that order, because the instrumental argument is the one that gets a hearing.
The default mode network
For most of the twentieth century, neuroscience treated the brain at rest as a brain doing nothing, a baseline against which interesting activity could be measured. In the late 1990s, Marcus Raichle and others noticed that the resting brain was not resting. A specific, distributed network lit up reliably the moment a task ended, and went quiet the moment a new one began. They called it the default mode network.
The default mode network consolidates memory, integrates experience, plans for the future, models other minds, generates autobiographical narrative, and does the slow integrative work of creative problem-solving. It connects things you did not know were connected. It is largely incompatible with focused, task-positive attention. The moment you pick up your phone, open a tab, start a podcast, or check Slack, you suppress it and switch to the task-positive network. The kind of cognitive work the default mode network is built for requires that you not do the other thing. Idleness is the only condition under which the network operates.
This is not a soft, motivational claim. It is a structural property of how the brain organises its work. The integrative thinking that distinguishes a good architect from a competent one, a founder who sees the shape of the market from one who reacts to the latest customer, a senior engineer who anticipates failure modes from one who debugs them after the fact, all of this is default-mode work. It cannot be scheduled inside a task. It happens between tasks, or it does not happen.
Rich Hickey gave this its best-known name in software: hammock-driven development. His talk is closer to a moral argument than a technical one, dressed in the language of system design. Hard problems are not solved at the keyboard. They are solved by holding the problem long enough, quietly enough, that the parts of your mind that do not work to deadlines have the chance to assemble it. The keyboard is for typing in the solution after it arrives. Most teams have inverted this. The keyboard is where they hope the solution will appear, and the hammock is treated as time stolen from real work.
The industry has produced quieter rediscoveries of the same idea under other names. Cal Newport’s deep work gestures at it but stays inside the productivity frame, treating idleness as a recovery period that exists to make the next sprint of focused work possible. Andrew Smart’s Autopilot argues the neuroscience more directly: that mental idleness is a precondition for the thinking we say we value. Jenny Odell’s How to Do Nothing argues the political case, that the attention economy is structurally hostile to the conditions any creative or integrative work requires. Each circles the same observation from a different angle. The integrative mind has a substrate. The substrate is idle time. There is no way to replace it with something that feels more legitimate.
You are not a productive unit
If the only argument for doing nothing were that it makes you more productive, the argument would collapse the moment it became inconvenient. The deeper case is that the framing under which idleness must justify itself, but how does it help me ship more, is itself the problem. The framing is industrial. It is a piece of nineteenth-century factory thinking that crawled out of the mill and into the office and now into the laptop, and it has been allowed to occupy the moral high ground in a knowledge economy where its premises stopped applying a long time ago.
Bertrand Russell wrote In Praise of Idleness in 1932, in the middle of the Depression, when the moral defence of work-as-virtue was at its loudest. His argument was simple. Modern productive capacity is high enough that the species could have its needs met with a four-hour day, with the rest of the time belonging to whatever each person found meaningful. The reason this has not happened is not technical but ideological. A culture has decided, against its own interest, that idleness is morally suspect and that human worth is measured in output. Russell saw this as a hangover from feudal and pre-industrial conditions when scarcity made constant labour rational, persisting into a world where scarcity no longer demanded it. Almost a century later, with productive capacity orders of magnitude higher, the ideology has hardened rather than softened.
Josef Pieper made a different version of the case in Leisure: The Basis of Culture. Leisure, in Pieper’s sense, is not the opposite of work. It is not a recovery period. It is the condition under which a human being can be receptive to ideas, to other people, to reality, rather than instrumentally pressing forward against it. A culture without leisure produces no philosophy, no art worth keeping, no science beyond the narrowly applied. It produces instead an unending sequence of better-tooled people doing badly thought-through things. Read in 2026 against the output of the average tech roadmap, the diagnosis lands with some force.
Pascal got there earliest and most economically. “All of humanity’s problems stem from man’s inability to sit quietly in a room alone.” Aphorism, not argument, but worth keeping in mind when you notice yourself reaching for the phone the second the silence becomes uncomfortable.
The point is not that productivity is bad. The point is that the assumption you are accountable, as a person, for being maximally productive at all times is a moral claim, not a technical one. It is a recent and parochial claim that does not survive contact with how thinking actually works or what a life is for. You have the right to exist as something other than a unit of output. The fact that you have to argue for this in a knowledge-work culture is itself the symptom.
What doing nothing actually means
The version of doing nothing I want to defend is specific, because the word has been hollowed out by adjacent practices that share its surface and miss its substance.
It is not meditation. Meditation is a structured attentional practice with goals, techniques, and progress markers. It is valuable. It is not what I am talking about. Meditation, in its modern productivity-coded form, is doing-nothing repackaged as another thing you can be good or bad at, with an app to track your streak. The whole point of the practice I am describing is that there is nothing to be good at.
It is not a walk with a podcast. It is not a drive with the radio on. It is not journaling, mind-mapping, or thinking time with a notebook. Each of these is a structured task that uses the surface of idleness as cover. The default mode network engages when the task-positive network disengages. Holding a podcast in your ear keeps the task-positive network online and the integrative work suppressed.
It is not relaxation in the wellness sense. It is often uncomfortable. The first few minutes feel like withdrawal, because that is what they are. A nervous system trained to produce dopamine in response to inputs is being asked to sit in their absence. The discomfort is the point. It is the sound of a habituated reflex going unrewarded. If you let it run its course, it subsides, and the condition the practice exists for arrives behind it.
It is sitting with the discomfort that screams at you to be productive, and refusing to obey it. Twenty minutes, an hour, half a day. Long enough that the productive-unit reflex exhausts itself and the rest of you can come into the room. No input. No output. No task. No measurement. Whatever happens, happens. Often what happens is that the problem you have been failing to solve for a week resolves itself in fifteen minutes, and you discover that the obstacle was never the problem; it was the conditions you were trying to think under.
This is hard to defend in a working culture because it is unfalsifiable from the outside. Someone sitting at their desk thinking about nothing is indistinguishable from someone slacking off, and the manager-class instinct is to assume the second. The discipline can be defended only by people who have done enough of it to know the difference, which is a small constituency.
What this means in software, startups, and product
The fields I work in have a particular relationship with idleness, and it is mostly hostile. Three observations.
Software architecture is hammock work. The cost of an architectural decision is not paid at the moment of the decision. It is paid over months and years, in the difficulty of every subsequent change. The decisions that matter, where the seams go, what is shared and what is isolated, what the unit of consistency is, where the async edges live, what the system refuses to do, are not the decisions a sprint planning meeting can produce. They are the decisions that arrive after you have held the shape of the system in your head, silently, long enough for the bad versions to fall away and the good version to assert itself. Teams that try to architect inside the sprint produce architectures shaped by the sprint: short-horizon, locally optimised, regretted at the next refactor. The fix is not better planning ceremony. It is making the conditions under which integrative thinking happens part of the working week.
Startups starve the founder of the only thinking that matters. A founder’s job is not to execute. Execution is the easy part once the question is right. The job is to keep asking what the question actually is, what the product is for, who it is for, which signal is real and which is the warm-network feedback I wrote about in When Things Slow to a Crawl. That work is integrative, slow, and incompatible with the calendar-shaped life most founders accept by default. The founder who is in twelve meetings a day and proud of it is, in practice, refusing to do the part of the job only the founder can do. Y Combinator’s standing advice to founders, fewer meetings, longer thinking blocks, write things down, points at the same observation from a different angle. It is the most under-followed advice in the industry.
Product is, at its core, a discipline of noticing. The best product decisions I have seen were not the output of frameworks. They were the output of someone who had sat with a problem long enough to see the shape of what users were actually doing, as distinct from what they were saying or what the dashboards were measuring. Noticing is default-mode work. A product team in continuous discovery mode, interview, synthesise, prototype, test, repeat, with the loop scheduled tight, produces motion that can look like rigour, but the synthesis step requires conditions the loop refuses to provide. Without it, the team converges on whatever the loudest input was, polished into a feature.
In all three cases, the failure mode is the same. Visible activity is substituted for the integrative thinking the visible activity is supposed to serve. It is the same failure mode I described as mistaking activity for progress in the antipatterns essay, viewed from the inside rather than the outside. The fix is the same. Some of the most important work a knowledge worker does looks, from a metre away, exactly like nothing. The work cannot be done under conditions where it is forbidden to look like nothing.
Treating it like exercise
The closest practical analogue is physical exercise. Most knowledge workers now accept, more or less, that thirty to sixty minutes of exercise several times a week is not a luxury that competes with work but an input that work depends on. The argument took decades to land, and it landed because the alternative, sedentary professionals quietly destroying their bodies in their forties, became visibly unacceptable. The case I am making for unstructured idleness is the same case, applied to the cognitive side of the same person, and not yet won.
The practical advice is the same as the practical advice for exercise.
Put it on the calendar. If it is not scheduled, it will be displaced by anything that is. A block on a Tuesday afternoon labelled “thinking”, or more plainly “nothing”, survives the week better than a vague intention to find time for it.
Defend the block. The first invitation to override it is the test of whether the practice will survive. The default outcome is that it will not. If the practice is treated as the thing that gives way every time something more urgent appears, the urgent will appear continuously, and the practice will never happen. The discipline is in saying no to the override.
Accept that the first sessions feel useless. They are not useless. They are the sessions in which the habituated reflex is being unwound. The integrative work begins after the reflex has gone quiet, which it does not do in the first five minutes. If you abandon the block the moment it gets uncomfortable, you have gone to the gym and left after the warm-up.
Do not try to extract a deliverable. The moment the block has to produce something, a decision, a document, a written-up insight, it has been recolonised by the task-positive frame, and the conditions you were trying to set up have collapsed. The insights, if they arrive, arrive on their own schedule, often hours or days later, in the shower or on a walk. The block’s job is to make their arrival possible, not to extract them on demand.
Do it more than feels reasonable. The right amount of unstructured idleness for a senior software career is probably several hours a week, not several minutes a day. This sounds extravagant only because the baseline assumption is that every working hour is owed to legible output. Once you let go of the assumption, the number stops looking extravagant and starts looking like the obvious cost of doing the integrative half of the job competently.
The paradox to manage
A working practice this easily mistaken for slacking has to be defended carefully against the version of itself that is slacking. The two are not the same thing, and conflating them is the fastest way to discredit the practice and to fail your team.
Begin with what slacking actually is. Slacking is almost never doing nothing. It is doing something else. Scrolling, watching, chatting, browsing, tab-switching, anything that floods the task-positive network with low-grade dopamine while the work that was supposed to happen does not. The practice I am describing is the opposite of this. Slacking holds the task-positive network online with low-value inputs. Doing nothing disengages it deliberately so the other network can run. They feel different from the inside, and they produce different results.
Work ethic still matters. The point of the practice is not to do less work. It is to do the part of the work that the calendar refuses to make room for, and to take it as seriously as the parts that get scheduled by default. The same person who blocks out two hours on a Tuesday to sit and think should still be the person who ships, who finishes what they started, who shows up to the meetings that matter, who holds themselves to the same standard of follow-through they would expect from anyone else. The practice is added to the discipline, not substituted for it.
Team coordination still matters more. A senior engineer who vanishes into thinking-time without warning, missing standups, unresponsive to messages, blocking other people’s work, has not adopted a practice. They have built an excuse. Other people’s work depends on knowing where you are and what you are doing. The discipline is in making the practice predictable enough that the team can plan around it, the same way a team plans around a colleague’s regular gym block, or therapy appointment, or school run. The block is on the shared calendar. The team knows it exists. Urgent things route around it. Non-urgent things wait for it to end. None of this requires explaining what happens inside the block, only that the block exists and that it is honoured.
Getting things done still matters. The practice is justified, in part, by the claim that it makes the work better. The claim has to be cashed in. If a quarter goes by and the work is no better, no clearer, no further along, the practice is failing on its own terms and needs adjustment, not defence. The integrative half of the job is half of the job, not the whole of it. The other half is shipping, and shipping is what reveals whether the thinking was any good.
The paradox to hold is that the practice both is and is not work. It is work in that it serves the work, requires discipline, takes a real share of the working week, and is owed to the team in the form of better decisions and clearer judgement. It is not work in that it cannot be measured, instrumentalised, or extracted from on demand without collapsing the conditions it depends on. The paradox is resolved not by resolving it but by holding both sides at once. You owe your team the integrative thinking only the conditions of idleness make possible. You also owe them the predictability, the follow-through, and the shipped work that idleness alone never produces. A senior career is built on doing both, and on knowing which one the current hour is for.
A short note on the AI moment
Worth saying briefly because it is the air everyone is breathing in 2026. The arrival of capable AI assistants has dramatically lowered the cost of the task-positive half of knowledge work. Code gets typed faster. Documents get drafted faster. The mechanical surface of the work compresses. This is, on balance, good. It also makes the integrative half of the work, the half that decides what to type, what to draft, what is worth building, proportionally more valuable, because it is now the half that produces almost all of the differentiated output. The professionals who win in this environment will not be the ones who execute the fastest with the most assistants. They will be the ones who have preserved the capacity to think clearly about what is worth doing, and the working conditions under which that capacity operates. Idleness, defended as a working practice, is one of those conditions. It has gone from a quietly useful habit to something close to a competitive advantage.
The whole point
Two arguments, restated plainly.
The first. The default mode network is the substrate of integrative thinking, and integrative thinking is most of what software, startup, and product work actually rewards. Idleness is the only condition under which the substrate operates. A working life that never allows the conditions makes the thinking it depends on impossible, and substitutes activity for it. This is not a soft observation. It is a structural one, and the science is settled enough that the burden of proof has moved.
The second, and the one I would rather lead with if the culture allowed it. You have the right to exist as something other than a productive unit. The framing under which idleness has to defend itself is a piece of inherited industrial thinking that survived past its usefulness and is now actively damaging the kind of work that has replaced industrial labour. You do not owe anyone the conversion of every waking hour into output. The right to a portion of your life that is not in service of anything is not a luxury you have to earn by being productive enough in the rest of it. It is a baseline condition of being a person, and a knowledge-work culture that has forgotten this is impoverished in a way it has not yet noticed.
Russell, Pieper, Pascal, Hickey, Odell, Smart. These are not the only voices saying this, but they are good ones, and they have been saying it for a long time. The case has been made. What is missing is the practice. Block out the time. Sit with the discomfort. Let the screaming subside. The thinking arrives behind it, and so does the rest of you.