Given the authors credentials, I would read all of the brain science in this book at the level of newspaper reporting - interesting, but wants expert confirmation. However the conclusions they draw from that science are, in my anecdotal experience, self-evident, so the book is still valuable.

Deep work is valuable

Deep work: Professional activities performed in a state of distraction-free concentration that push your cognitive capabilities to their limit. These efforts create new value, improve your skill, and are hard to replicate.

Shallow work: Non-cognitively demanding, logistical-style tasks often performed while distracted. These efforts tend to not create much new value in the world and are easy to replicate.

Digital lifestyle tends to push towards shallow work/interaction - social media, instant connectivity, infinite supply of infotainment.

But new technologies also increase value of deep work. Firstly, specialization and rate of change have both increased, which means that the ability to quickly learn complex subjects is increasingly valuable. Secondly, easy distribution provides leverage, increasing the value of the best work and reducing the competitiveness of everything but the best.

Increasing division between high skill jobs and everything else, due to distribution, automation and globalisation. ‘Average is over’, winner takes all. This ‘great restructuring’ favors:

Deliberate practice as the foundation of expert ability. Requires that:

Not compatible with persistent distractions and diffused attention.

At the neurological level, learning requires repeatedly isolating circuits. Distracted attention fires too many different circuits.

Task residue - in serial single-tasking, performance on each task is affected by the previous task. This residue is worse if the previous task was unbounded, low intensity or left uncomplete. Context switches in humans are costlier than we are consciously aware of. Implies that eg checking emails while your code compiles is not a pleasant break, but actually inhibits performance for long periods afterwards.

The type of work that optimises your performance is deep work. Slightly fuzzy arguments in places, but I think it’s driving at focusing on increasing comparative advantage - being able to do some particular thing far better than the average person.

This only applies to some kinds of work eg CEOs are necessarily engaged in shallow work most of the time because their job is to act as a decision hub. The ideas in this book do not make sense if your job requires constant, diffuse attention. ‘Shallow work’ might hit pushback because of the connotations. Would make more sense to contrast focused attention vs diffuse attention, but that would be a less catchy title.

Several anecdotes about people who believed their work fell under this category but who found that when forced to disconnect regularly their value actually improved. This suggests experimenting with different ratios of time spent on focused vs diffuse attention, to determine what balance works best for your situation.

Deep work is rare

Open plan offices. Company chat rooms. Social media presences as part of professional responsibilities.

If deep work is so valuable, why the move towards distraction?

Metric black hole - can’t easily measure productivity for most knowledge workers. Insecurity results in pressure towards appearance of productivity instead - look busy, send lots of emails etc.

Cochran experiment measuring time spent on email, resulting in company-wide guidance on how to reduce email to only the most valuable.

Consulting group forced to spend one day per week without email access. Resulted in better satisfaction all round.

Knowledge workers generally don’t have a clear metric by which to judge their work, and even when they do it’s usually a very long causal chain away from their actual actions.

Principle of Least Resistance: in a business setting, without clear feedback on the impact of various behaviors to the bottom line, we will tend towards behaviors that are easiest in the moment.

Planning and prioritizing is hard. Responding to the top email in your inbox is easy. I notice that many people, myself included, tend to start the day by opening their email and then often don’t make it to actual work until lunch.

Same goes for excessive status meetings as a form of deadline / motivation control, and to provide constant reassurance of progress.

Metric black hole + principle of least resistance + busyness as proxy for productivity = shallow work.

I find myself in the weird position of believing the claims to be roughly true, but still not being convinced by the argument as presented in the book.

Cult of the Internet. Technopoly. Assumption that anything high-tech is inherently good, and is the future. Refusal to adopt seen as Luddism, couldn’t possibly be the result of consciously weighing costs and benefits.

Deep work is meaningful

Attention determines experience, as much as circumstances. Some links between mindfulness and happiness / resistance to depression. Also some evidence that executive function can be increased with practice, with far transfer. With intense enough focus, there is not enough leftover attention for anxiety or worry.

Shallow work is draining, unsatisfying, stressful. Evidence provided is slim, but anyone who is reading this can probably conjure up a contrast between memories of good work days vs memories of days spent refreshing inboxes. Fairly confident that experience sampling would confirm that diffuse attention at least correlates with stress.

Flow experiments - experience sampling finds that “best moments usually occur when a person’s body or mind is stretched to its limits in a voluntary effort to accomplish something difficult and worthwhile.” Huh, well there you go.

Combat nihilism by finding meaning in quality work eg programmers finding aesthetic satisfaction in code. Shades of Zen and the Art of Motorcycle Maintenance here, except that this book isn’t a muddy mess :P

It’s as if our species has evolved into one that flourishes in depth and wallows in shallowness.

I notice that everything in this chapter seems obvious to me, and yet I still behave as if internet time is a reward for the punishment of hard work, even though I know I generally enjoy the latter more. How did this idea worm it’s way into my head?

Work deeply

Willpower is a depletable resource - can’t rely on it to avoid distraction. Instead cultivate habits / rituals / routines around deep work and design your environment to remove distraction.

1) Decide on your depth philosophy:

2) Ritualize. Plan out the structure of your deep work periods - where you will work, for how long, how you will work, what you will work on, how to remove distracting obstacles ahead of time.

3) Make grand gestures. Can be big, like buying a round-the-world flight to force yourself to disconnect. Can be small, like buying a fancy notebook to plan your schedule, making it seem more important.

4) Don’t work alone. Collaborative work in a good environment can be deep, and a hub-and-spoke -style office can support serendipity without constant distraction.

5) Execute like a business. Use the 4DX framework:

6) Be lazy. Downtime appears to be important for both creativity and concentration. Good downtime is playing with your kids, taking a walk in nature, going for a run. Bad downtime is browsing facebook etc. Not clear what the dividing line is - perhaps something like mindfulness/presentness vs compulsive consumption.

Create a shutdown ritual. The authors is: check email one last time, dump brain onto todo list, skim calendar for important deadlines, plan the next day, announce out loud “shutdown complete”. After the shutdown, NO WORKING! Not even checking email.

Embrace boredom

Focus is trainable, and conversely distraction is an addiction. If you pull out your phone every time you have a spare minute, you are feeding the addiction and missing the chance to practice focus.

Constant stimulus/distraction causes long-term changes in the brain that prohibit the ability to focus deeply.

Practice resisting distraction. Divide the day into focused and distracted periods, and enforce that separation. Even if your job requires eg regular email responses, you can still divide the day into 20 minutes offline and 20 minutes online. The important part seems to be installing a check between feeling the urge of distraction and acting on it - so that instead of just finding yourself mysteriously on facebook you first have to consciously decide if it is currently allowed.

It’s not shallow interaction itself that is dangerous, it’s the constant switching back and forth. Need to learn to tolerate an absence of novelty.

I once spent a week in Budapest with a good friend where I forgot to bring my phone charger. As a result I only turned my phone on when we were totally lost and needed the map. I was surprised by the end of the week how much that affected my experience. I felt strangely calm and relaxed.

“Interval training for the brain”. Pick a deep task and give yourself a challenging deadline to complete it. Provides an immediate motivation to avoid distraction.

A few months ago I decided to dedicate a week to reading. I set myself what I thought was a steep challenge - to read and publish notes on five books. I ended up finishing nine, whilst only spending 6-7 hours per day on the task. The fact that I was able to focus for that long was down to an unusual environment and motivation, but it did make me wonder what I could do to replicate that in general.

Since then, I’ve toyed with the practice of setting extreme deadlines on small programming projects eg one hour. This has two interesting effects. The first is that it drives one towards just solving the core of the problem and pushes out any busywork that doesn’t contribute to the solution. The second is that it really highlights and inefficiencies or flow-breaking in the tools. Normally a 3-minute compile is just a chance to check my email, but when I only have 60 minutes to begin with it’s suddenly a massive imposition.

Meditate productively - during time that is otherwise physically occupied (eg walking, commuting) set yourself a mental task to attack. Like mindfulness meditation, notice when you have drifted away from the task and gently pull yourself back. Avoid looping around the same thoughts by noticing and banning the loop. Structure your process to avoid stalls eg begin with a review of the relevant variables, list the next step questions and then attack each question one by one.

Memory training - appears to actually train concentration as much as memory. Any process which requires unwavering concentration might have the same effect. In a similar vein, I’ve been using rail balancing to practice mindfulness. It provides a very clear signal - if my mind wanders, I fall off.

Quit social media

Social media is explicitly designed to be addictive. The metric that companies like Facebook and Youtube optimize for is time spent on the site. See Is Anything Worth Maximizing?_

But part of what makes social media insidious is that the companies that profit from your attention have succeeded with a masterful marketing coup: convincing our culture that if you don’t use their products you might miss out.

The any-benefit approach to network tool selection: You’re justified in using a network tool if you can identify any possible benefit to its use, or anything you might possible miss out on if you don’t use it.

The craftsman approach to toll selection: Identify the core factors that determine success and happiness in your professional and personal life. Adopt a tool only if its positive impacts on those factors substantially outweigh its negative impacts.

In other words, consider carefully the opportunity cost, especially on your ability to focus.

Note that just like Decisive it encourages explicitly choosing your values and then using them to weigh choices about how you spend your time.

The core factors should be non-completable, but still specific enough to be actionable. ‘Do better research’ is too vague. ‘Finish paper for X conference’ is too completable. ‘Regularly read and understand cutting-edge results in my field’ is good.

Can be hard to tell what impact social media has on your life, so determine it experimentally - go cold turkey for one month and keep track of the answers to these two questions:

Don’t use the internet to entertain yourself (at least, don’t use it unboundedly). Plan your free time in advance to ensure you make good use of it, instead of falling into low-value time sinks like social media.

“The great and profound mistake which my typical man makes in regard to his day”, he elaborates, is that even though he doesn’t particularly enjoy his work (seeing it as something to “get through”), “he persists in looking upon those hours from ten to six as ‘the day’, to which the ten hours preceding them and the six hours following them are nothing but a prologue and epilogue.”

This is a perfect opportunity to plug Microadventures. Most people think they don’t have time for interesting adventures because they are busy working 9-5. Alastair turns that around and points out that you have all of 5-9 to do something memorable.

Drain the shallows

Deep work is tiring. You may only be able to manage four hours per day. But shallow work is surprisingly pervasive and, combined with task residue, can easily crowd out those four hours.

Put a notebook on your desk with an hour-by-hour schedule. Fill the whole schedule at the start of the day, and adjust it as needed during the day. You don’t have to rigidly stick to the plan eg if inspiration strikes feel free to spend the rest of the day mining it, other plans be damnded. The idea here seems to be a combination of intentionality about your time and creating a more explicit awareness of where your time goes. I like the idea of planning the whole day in advance more than the short-term planning I have built into monolog at the moment.

Quantify depth of activities by asking how many months would it take to train a smart college grad to do this job. Then bias your time towards activities where you provide more value than a smart college grad! This is intended as a thought exercise but, being self-employed, paying a smart college grad to take over some of my work is actually an option.

Ask your boss what percentage of your time should be spent on shallow work, and then use this to fight back against shallow demands on your time. Your boss is unlikely to commit out loud to the idea that, say, 80% of your time should be spent on shallow work (if they do, start looking for another job). This is genius - weaponizing the consistency effect.

Fixed-schedule productivity - rather than working until the work is done, pick a schedule and make the work fit into the time available. Like setting a deadline for individual projects, this has the effect of squeezing out non-essential activities and motivating focus throughout the day. Also motivates guarding your time more preciously - treating it as a highly limited resource rather than saying yes to every request that comes to you.

Become hard to reach:

These suggestions probably make more sense if you are a famous author and professor with a huge inbox. Personally, I’ve found the opposite effect - that I get a lot of value out of explicitly encouraging people to contact me. I do often ignore thoughtless emails though eg I regularly get inquires about web dev projects, which clearly indicates that the sender didn’t so much as glance at my website or portfolio before contacting me.


I think the most valuable thing I got from this was having the inconsistency pointed out between my experience and my actions - I know that deep work is what I find satisfying, but I motivate myself as if random internet browsing is a reward for the punishment of work.

I used to make a daily schedule as the book suggests. It’s a habit that got lost when I spent a few years in a regular job and didn’t have the power to reject interruptions. Monolog currently has a task tracker that requires to state what I am working on and for how long, and then nags me if I go over time. I think that planning the whole day in advance is even more valuable, because it highlights the opportunity costs in spending too much time on less valuable activities.

A daily plan is also the most actionable idea here. It would be very easy for me to agree that all of these things are good ideas, but not get around to making them into habits. Planning the whole day in advance provides a perfect nudge to consider including all these other ideas.

Explicitly writing down personal values and then mapping them to activities is a suggestion I’ve seen in a few places now. I have some cute ideas about software that encourages managing time well according to your stated values, and makes it clear if some value is being neglected, but for now I’m just going to use paper and a one hour timer.

Deadline tasks are something I’ve already been doing occasionally, but casting it as ‘interval training for the brain’ makes me even more bullish on the idea.

The concepts of fixed-schedule productivity and shutdown rituals are well-timed, because I’ve just started consulting again and in the past I always struggled with switching off from work.

I want to set up a proper office but I’m currently constrained by living situation. I’m looking at local coworking spaces but I don’t think it’s likely that any of them will have private rooms.

Speaking of rituals, I want to buy a thinking hat. The rule will be that I can answer email or chat online or whatever I want, but first I have to take off the thinking hat. I don’t get to pretend that this is still included in deep work time. I’ve found little speedbumps like this, that place conscious action between impulse and gratification, are really valuable for creating awareness of my own behavior.

“I’ll live the focused life, because it’s the best kind there is.”