Ideas about how to effectively manage your brain, based on modern science.

Author is an executive coach rather than a researcher, but has been highly recommended by several researchers. Still, tread with caution.

Each chapter starts with a story illustrating a poorly handled situation. Explains the mechanisms at play. Then shows how the story could have gone better by following advice. I felt like these stories helped make the ideas stick with me better so they are worth reading, but I won’t attempt to summarize them here.



Goal is to work smarter, be more focused and productive, stay cool under pressure, reduce the lengths of meetings and influence other people.

Problems and decisions

The morning email overwhelm

Prefrontal cortex. In evolutionary history, last major brain region to develop. Involved in setting goals, controlling impulses, imagining fictive scenes. “It’s where we hold thoughts that are not being generated from external sources or from the senses.”

Limited computation power. Prefrontal cortex is to rest of brain as change in your pocket to the US economy. Goldilocks - everything has to be just right for it to function well.

Metaphor: Prefrontal cortex is a stage. Actors are information held in your attention. Sometimes they enter from the side - info from the outside world. Sometimes the audience jumps on stage - ideas, knowledge, thoughts and memories from your own mind.

To understand an idea, you put it on the stage and hold it there long enough to see how it connects to audience members.

To make a decision, you hold multiple actors on stage and compare them. To make a conscious decision? Compare to research on unconscious decision making, which doesn’t require attention and can better compare choices with many facets.

To recall information, you call it up from the audience. If it’s dimly remembered, it may be in the dark at the back of the crowd and might take time and effort to find. What about when you fail to recall something and hours later it just pops into your mind. Does recall actually require conscious attention?

To memorize information, you need to get if off the stage and into the audience.

Inhibiting a thought - trying to stop an over-excited audience member from getting onto the stage - takes effort but is essential to saving the limited capacity of the stage.

Effort = oxygen, glucose. Prefrontal cortex is a pig - can measure drop in blood sugar when active. Will shut off when out of resources. This is why you are easily distracted when tired or hungry.

(Basal ganglia - control routine actions. Highly efficient.)

Implication is that attention/focus should be treated as a limited resource, to be deployed where it is most valuable. Don’t give attention to things that are unimportant/ineffective. Be disciplined in not thinking about problems that you can’t currently solve. Personal habit is to reflexively check whether I can do anything right now, and if not then set a reminder to think about it at some future date. The existence of the reminder stops me worrying about it.

Prioritization requires prefrontal cortex. Don’t allow any distractions (eg email) before prioritizing your day. Parallels popular advice to start your day by doing the most important thing.

Recall of chronological events (eg what was for lunch 10 days ago) requires walking backwards in time. Recall difficulty is proportional to time since event.

Imagining something unfamiliar takes more effort than imagining something familiar. Easier to think about current problems than possible solutions.

Have very efficient circuitry for social interactions, images and narratives. If you can convert an abstract problem into a social/visual/narrative problem you get to take advantage of this. Is this why metaphors are useful?.

Practices like writing down thoughts/worries/ideas can aid inhibition, freeing up attention for more useful work.

Schedule tasks that require attention at times when you expect to be fresh.

A project that hurts to think about

Stage is small. Short-term memory seems to be around 4 items. Can maybe only remember 1 item without any degradation. Figuring out what constitutes an item is tricky, probably depends on what supporting mental structures you have - could be 4 digits or 4 words or 4 sentences, if the sentences make sense.

Relational complexity - when making decisions, the fewer variables you hold in mind at once, the more effective you are at making decisions.

Competition between circuits. Contradictory circuits inhibit each other. Can only focus on one at a time, but can switch between them at will.

Simplify - reduce ideas to a size that can fit into focus.

Chunking - creating mental structure to aid recall eg grouping phone-numbers into 3-4 digits. Chunks seem to work best when they take about 2s to think or say.

Part of becoming an expert at something seems to be in creating new kinds of chunks, allowing to think more complex thoughts without changing the limits of the prefrontal cortex.

Prioritize what gets focus. Use external storage (eg writing) to offload chunks that are not immediately needed.

Interesting to relate this to previous notes on Fermi equations or cost-benefit analysis. Is the main benefit of these externalizing the structure so that we apply focus to each detail of the problem in turn?

Juggling five things at once

Another limitation - stage can only perform one process at a time. Processes include understanding, making a decision, memorizing, inhibiting.

Can only multitask effectively when all but one of the processes do not require conscious attention. Dual-task interference - context switching between two mental tasks tanks performance.

Constant texting and emailing has same effect.

Allostatic load - measure of stress hormones and other factors related to sense of threat - physiologically damaging. Being ‘always on’ leads to brain being stuck on alert.

Does this imply that I shouldn’t switch too much between note-taking and reading? Had been considering switching to chapter-at-a-time. Will try that on the next book. The effort to remember the contents of a chapter long enough to take notes may be useful in itself.

Solutions to multitasking problems:

More automation (mental or technological). Automatic processes don’t steal focus. Eg learn to touch-type, learn keyboard shortcuts. Develop routines for daily tasks so that there are patterns for the basal ganglia to learn. The combination of fullscreen book, zen mode editor and well memorized keyboard shortcuts to switch between them has been very useful for writing these notes.

Handle recurring mental distractions. Solve nagging problems. Find ways to address or delegate worries. Don’t let unmade decisions sit in your mental queue.

Explicitly partition attention. Schedule times to check email. If interrupted by a call, abandon current task rather than trying to do both.

TODO There are lots of dramatic physiological signs that conscious effort is being applied. Can I find a way to continuously measure this on myself? Can I detect exhaustion and warn myself to stop?

Saying no to distractions

Distractions, both external and internal, require effort to inhibit.

Most common distraction I notice right now is going from understanding something to thinking about how I would explain it to someone and then drifting away into thoughts of eg how to structure a talk.

Seems likely that mindfulness practice would be effective in reducing internal distractions since it increases awareness of conscious thought.

Anterior cingulated cortex - involved in detecting novelty + errors. Small amounts of novelty are pleasant. Large amounts cause stress/alert/anxiety/fear.

Ventrolateral prefrontal cortex - involved in ‘braking’ / inhibiting responses - overriding instinctive reactions and replacing them with conscious, reasoned reactions. Requires energy, works less well when you lose your focus. Self-control is a limited resource. I note that the most likely time for me to break down and binge on icecream or chocolate is during a stressful work day. I just lose the ability to override the desire.

For subconscious decisions, there is a short (~0.3ms) interval between activation in brain and conscious awareness of decision. There is another short interval (~0.2ms) between conscious awareness and actual action. This is the window in which you can inhibit your response. Once you start the action and begin getting rewards (eg novelty reward for checking email) it’s much harder to control yourself. See also the focus in rationality of teaching 5-second level skills. Knowing principles is not enough, have to be able to reflexively apply them.

Don’t have free will (ie no control or introspection on subconscious decisions) but have “free won’t” - can inhibit action.

Easier to veto actions/thoughts if you have conscious labels for them. Can improve vetoing with practice. This last point doesn’t have any references yet? If true, does it transfer across different situations?

Searching for the zone of peak performance

Over-arousal reduces performance. There is a sweet spot between too much pressure/stress and not enough, where performance is maximized. Does this only apply to conscious tasks?

Norepinephrine - brain, feeling of alert/fear. Can increase alertness by visualizing scary consequences of failure. Don’t overdo it. In climbing, common advice is not to allow visualizing failure at all after initial preparation. Maybe just visualizing a hard but successful climb is enough to increase alertness? Does alertness even matter for non-conscious tasks?

Dopamine - brain, feeling of interest/novelty. Can increase by change of environment or by visualizing future rewards. Seems to be more effective than using alert/feat.

Bit confused about the relationship between the specific chemicals and the practices. Mentions that reward can generate both dopamine and adrenaline, which presumably increases norepinephrine.

Over-arousal - experienced as stress, panic, over-excitement, inability to think/focus. Reduce arousal by eg taking a walk.

Getting past a roadblock

Creative block - can’t generate novel ideas, keep circling back to some existing idea. Same actor keeps hogging the stage. Can’t think out of the box.

Need to reduce activation of the blocking circuit before others can be activated. If inhibition isn’t working, must switch off entirely. Even a very short break to a completely different activity can help.

Right anterior temporal lobe - involved with connecting distantly related information. Shows more activity when successfully solving a problem. During the process of solving the problem, presumably?

Insight is more likely in happy, relaxed states than in alert, focused states.

ARIA model - Awareness, Reflection, Insight, Action - both describes the stages of insight, and advises how to aid insight.

Awareness - viewing the problem we are stuck on. Want to leave space on stage for other ideas, so simplify problem as much as possible - elevator pitch. Don’t try to focus too hard.

Relection - reflect on thinking processes. Notice what strategies have been tried and what might be missing. Aiming for relaxed daydream-like state.

Insight - solution is discovered. Visible as burst of gamma waves in brain - lots of communication between different brain areas. Burst of dopamine and adrenaline.

Action - take advantage of short-lived high - extra courage and energy.

Meet the director

Executive function - director is the part of you that observes your mental processes.

This is the part of you can debug and rewrite your mental processes!

Interoception / mindfulness - awareness of your own thoughts and thought processes. Can be improved by training. Measured by Mindful Awareness Attention Scale.

Default network - involved in planning, daydreaming, ruminating. Also narratives. Called default circuit because it tends to kick in when not doing anything else.

Direct experience - focus on immediate sensation and perception, including interoception. Confused by this - my anecdotal experience is that direct perception and introspection are different states and are mutually exclusive.

Default network and direct experience inhibit each other. Mindfulness training increases the ability to notice and control the switch.

Stay cool under pressure

Derailed by drama

Limbic system - track emotional relationship to thoughts, events, people, objects. Executes value judgments.

Toward/away response - limbic system motivates you towards primary rewards (food, money, sex etc) and away from primary threats (predators, hunger, thirst etc).

Not very bright - activates on eg scary pictures or bad memories.

Emotional arousal often measured by activation of amydale.

Walk toward, run away - away responses trigger faster, last longer and more intense. Towards responses are more easily displaced. This is why negative spirals are more common than positive spirals.

Over-arousal of limbic system can impair decision-making. Can also sometimes increase confidence (eg adrenaline rush). Poor combination.

Surprisingly strong effect. Triggers as simple as smiley/frowning faces can impact performance.

Over-aroused limbic system also makes more sensitive to false pattern-recognition. When you are scared, every tree is a monster.

Attentional blink - interval between identify different stimuli. If internal stimuli are competing with external stimuli, get more blinks and can miss info.

Before emotional arousal: Situation selection - avoid situations that cause high emotions. Situation modification - do what you can to make the situation less impactful eg prepare thoroughly. Attention deployment - control thoughts to avoid spiralling, focus on other stimuli.

During emotional arousal: Express - eg cry - often not an option but worth considering when it is. Expressive suppression - hold it in - diverts cognitive resources. Cognitive change - use conscious control to influence limbic system eg via labeling or reappraisal.

Labeling - simply attach a label to the emotion eg “I notice anger”. Seems to inhibit limbic system. Use simple label, not lengthy description, to minimise focus given to the emotion.

Higher scores on MAAS correlate with better effectiveness of labeling. Note that common to many mindfulness practices is labelling and then releasing thoughts/emotions.

Drowning amid uncertainty

Prediction is one of the core functions of the brain.

The limbic system is activated towards certainty and away from uncertainty. Strong response - uncertainty is treated as a threat.

Similarly for autonomy/control. Feeling of control, the ability to take action, can reduce the threat from uncertainty.

Reappraisal, reframing, re-contextualising - finding a different perspective on the situation. Cannot directly control limbic system, but can influence it by looking for perspectives which generate the desired level of emotion.

Types of reappraisal. Reinterpreting - new information brings an new understanding - “oh, they’re crying because they’re happy about getting married”. Normalizing - recasting the current situation as expected - “it’s normal to feel overwhelmed when starting a new job”. Reordering - place focus on different values - “getting fired is actually exciting, it means I now have freedom to choose a new life”. Repositioning - deliberately simulate a perspective other than your own - “if I were in their shoes I would be angry too”.

People who use reappraisal rather than suppression are better off on a variety of measures of emotional health such as life satisfaction.

Discusses whether reappraisal is tantamount to lying to yourself. Note that Tetlock’s foxes made more accurate predictions by synthesizing many perspectives. It’s totally plausible that reappraisal could lead to more accurate perceptions on average.

Suggests using reappraisal when frustrated with yourself too - instead of “I screwed up” use “my brain screwed up”, leaving yourself more resourcs to actually fix the problem.

When expectations get out of control

Expectations can influence your perceptions and reactions. Perceive patterns faster when expecting to see them. Also more likely to have false positive.

Unmet expectations can cause a threat response, as if you already had the reward and it was taken away. Seems similar to loss aversion - moving the reference point changes the value function. Unexpected rewards are more pleasant than expected rewards (in the moment, presumably. You don’t get the pleasure of anticipation with surprise rewards).

Placebo effect can reduce pain. Effect remains, though less strong, even when told it’s a placebo. (Can measure brain with brain scans, to avoid problems with self-reporting.)

Advises managing expectations, especially when you notice that they are causing strong arousal. If you need the prospect of a reward, focus on a certain reward rather than one that might not appear. I find it useful to focus on the satisfaction of a good days work as a reward rather than on the results, otherwise I can get demotivated when the results don’t match my expectations.

Collaborate with others

Turning enemies into friends

We also experience away/towards response for people. Classification into friend/foe is immediate and subconscious, and usually defaults to foe to be on the safe side.

The quality and quantity of your social connections is the one thing that has been empirically demonstrated to raise your happiness (once base needs are met).

Loneliness or feeling left out generates a threat response.

Speaking out loud while learning increases retention.

The foe response reduces empathy. Makes it harder to see the foes perspective. Makes you more likely to interpret intent as hostile. Contaminates your opinion of ideas that they support.

Defusing the stranger-danger can be as simple as sharing a joke or shaking hands. Banter at the start of a meeting or events that unite distant colleagues are examples of practices that can switch the categorization.

When everything seems unfair

Perceived fairness generates a reward and perceived unfairness generates a threat response. Perceptions of unfairness can be poisonous to any group and need to be addressed urgently.

Fairness and relatedness are linked. When people perceive that they are being treated fairly, they are less likely to perceive the other as a foe.

In game experiments, stronger activation of VLPFC is correlated with likelihood of accepting unfair offers (instead of rage-quitting and getting nothing).

You can generate fairness rewards for yourself by volunteering or helping.

The battle for status

Social status is another primary threat/reward. High social status is correlated with longer life (controlling for education and income) .

Tend to avoid situations that could threaten our status.

Social pain (exclusion, rejection etc) activate same regions of the brain as physical pain. Sticks-and-stones is empirically untrue.

Can game your perception of status by directing focus to an area where you feel better than others.

Giving positive feedback to others increases their feeling of status, leading to fewer destructive threat responses when collaborating. But if can feel like a threat to your own status.

Can also game status by ‘playing against yourself’ - gain status by being better than your previous selves.

SCARF - status, certainty, autonomy, relatedness, fairness. Social perceptions that are treated as primary rewards/threats, on a par with more direct threats to survival.

Be wary of creating threats to other peoples SCARFs - will lead to irrational responses.

Facilitate change

When other people lose the plot

Feedback is tricky. Most popular techniques ignore psychological findings like SCARF. “Let me tell you what I/others think about you” is a sure way to generate a threat response.

When you have a problem with someone, it’s natural to want to focus on talking about solutions. This threatens their autonomy and status - feels like telling them what to do. When feedback is often a threat, people learn to avoid giving/receiving feedback. Communication suffers.

Instead, help them come to a solution themselves by putting them into a good mental state. Increase status with encouragement, increase certainty by explicitly stating issues, increase autonomy by stressing that they are the person to figure out the solution. Encourage them to state the problem as simply as possible to reduce load on PFC. Ask questions which focus attention on their thinking processes rather than the problem itself. This is remarkably similar to the teaching methods described in How to solve it. The idea is that math teachers are responsible for teaching not just knowledge, but problem solving strategies, and that they must be careful how they help the student. Eg ‘the theorem starts with P’ does not demonstrate a useful strategy, whereas ‘what theorems can you think of that relate to this problem’ does.

In general, when solving problems, focus on the desired outcome rather the problem itself. Creates templates for pattern-matching circuits to look for. Avoids emotional over-arousal from the problem.

Natural to protect your status by defending your own work, missing the opportunity to learn. Managing the perception of status threat is crucial to growth.

The culture that needs to transform

Carrot-and-stick doesn’t work well with adults - sensitive to manipulation, resentful of attempted control, punishment creates threat response.

Don’t try to influence people during away state. Need to reduce threat response first.

Need to direct focus. Carrot-and-stick directs focus to external punishment/reward. Instead try to direct focus to internal motivations.

Instead of explicit rewards, appeal to SCARF. Eg rather than declaring punishments for a problem, tell people about the problem and give them power to fix it - appeals to status and autonomy.

Avoid uncertainty by explicitly bounding discussions beforehand eg ‘I only want to talk for 15 minutes and Im not looking for specific outcomes.’

Focus attention on the desired area by asking questions instead of demanding improvement eg ‘What is one thing you have done that has made a customer delighted in the past?’. People arrive at their own answers - no threat to status / autonomy. Draws focus away from negativity which could arouse emotional responses.

Be as clear as possible about the desired goal. Head off potential distractions. Eg focus on solution to problem rather than who is at fault.

Need regular attention to maintain new behavior. Set up collaborative processes that involve regularly talking about the behavior eg have a weekly brainstorming session on reducing customer complaints.


Even aside from the content, the book is really interesting in the ways it attempts to make the information actionable. Examples are placed in a familiar narrative rather than an artificial experiment, making it easier for the reader to recall them at the appropriate time. Each chapter ends with a short list of points to remember and actions to try in future (I found that the summaries in eg Mindware were too long to be recalled and mixed information with suggestions - I need chunked information). The book introduces a few core metaphors that everything else hangs off, providing a pre-made mental structure to organize all this new information around.

The content is much less rigorous than some other books and adds a lot of interpretation to the actual evidence. That may actually be a good thing in this context - for research it is important to be very clear about exactly what is known and what is assumed, but for teaching skills it’s more important to present a simple, cohesive model. See also ‘all models are wrong, but some are useful’. Given the book is aimed at teaching skills it should be judged on how effective it is rather than how accurate it is. So how to evaluate effectiveness…

Things for me to remember: