Make behavioral automation work for you instead of on you

You do a lot of what you do without ever thinking about it.

With a little reflection you’ll notice that you are not always sure what leads you to do some of the things that you do. Most of us have unconscious habits like a nervous laugh or finger tapping. Then there are those extra drinks or pieces of cake that we intend to refuse but end up (thoroughly) enjoying and the comments that burst out of our mouths despite our intention to hold our peace this time around.

Setting aside things you do that you didn’t intend to do, how good are you at doing the things you intend to do?  If you are like me, sometimes you do and sometimes you don’t. I stand up intending to make a cup of coffee but then I remember 20 minutes later that the beans are still where I left them in the grinder. I sit down with the intention to focus for 3-4 hours on an important project only to notice 90 minutes later that I am still clearing off email or reading the news. I start a tennis match intending to practice my topspin backhand but once the points begin all that seems to matter is winning. It sometimes feels like someone else is in charge of what I do and I am just there to watch the show.

Between intention and action comes automation.

As it turns out, there is a reason why we sometimes feel like we are the audience rather than the players in our own lives. Much of our behavior is to some degree automated. We are equipped with a network of physiological, mental and emotional circuits that enable us to learn from each situation so that we can respond more efficiently and effectively to analogous situations in the future. The automation process works to our advantage in many ways, but it also raises the question of whether we can drive it or if it is our fate to be driven by it.

Not all automation is bad.

For the most part, the brain’s automation works to your advantage. It enables you to execute complex physical feats such as running and talking without consciously thinking about either task or driving a car while thinking about how you will solve a problem at work. Automation enables athletes, dancers, musicians and even great orators to execute amazing feats without paying attention to the mechanics of how they do it. Behavioral automation is a gift that enables you to focus on the few actions that really require your attention while entrusting the rest of what you do to that crazy genius hiding in the networks that connect your synapses and sinews.

Automation doesn’t always get it right.

Unfortunately, automation doesn’t always work as we would like it to.  Sometimes we automatically do something we would rather not do. We make a turn that takes us to an old destination when we are actually supposed to be heading toward a new one.  We say things in the heat of the moment that we know are likely to harm our relationships. We make predictably irrational decisions, develop phobias and succumb to cognitive biases that nudge us in undesired directions.

It can be difficult to undo automated behavior once it has been installed.

When it comes to changing behavior we often know what we want to do, but we aren’t able to do it properly or even remember to do it. Routine situations lead to routine reactions. Like Pavlov’s dogs after a few rounds of exposure to a set of conditions or cues, we settle into a pattern of conditioned responses. What we actually do in real time is often driven by a mixture of environmental cues and physiological reactions that we may not even be aware of. With enough repetition, we go into autopilot, doing what we have done in the past without considering whether or not that is what we want to do now.

Experience is the great programmer.

The drivers of automation are experience and association. Repeated experience leads to formation of associations and those associations in turn become cues that solicit automated reactions. At some point we have experienced enough repetitions of the cue-association-reaction cycle to install an automated action. Since repetition plays a large role in automation, it is not surprising that much of the literature on the making and breaking of habits focuses on repetition as a key driver of change, but there is a catch…

It is hard to repeat what you can’t remember.

You can practice a new action or skill over and over again, but even after extensive practice you may not be fully aware on a detailed level of all of the elements of the action you want to take or the habit you want to break. Moreover, in real time, intention is not the only driver of your actions. The world is full of shiny objects that pull the mind away from its intentions.  Once you have been distracted you often find out after the fact that you have once again repeated one of those bad habits that you intended to break.

If you can’t consciously and directly control your habits, what can you control?

This is a topic that has received a lot of attention from neuroscientists, psychologists and philosophers. It touches on the essence of free will.  While the science on exactly how much of our behavior is automated and what we should do about it is not complete, patterns have started to emerge and a number of authors have suggested concepts and images that can be put to good use in the management of automation.

One line of thinking I find particularly interesting is the reconsideration of what it means to be a “self.” We naturally assume we each have a single, discreet unitary self, but research on the drivers of behavior indicates that this may not be the an accurate representation of the self. Some authors suggest that there might be more than one “you” in “you” and that your actions emerge through the interactions among those “yous.”

The self that thinks about action is not the self that takes action.

In Thinking Fast and Slow Daniel Kahneman suggests that the self breaks into two parallel systems: an automated self that acts without thinking and a conscious self that thinks about the implications of acting. The automated self, which is in charge most of the time, reacts to situations based on the default settings we have installed through past experience. As long as we are in familiar environments with familiar people doing familiar things, the automated self does a pretty good job of getting us safely through the day.

Sometimes we run into something that deviates from the patterns we have become accustomed to in our routinized lives. This deviation from expectations activates the conscious self so that it can pay closer attention to the situation, analyze it and choose the best actions based on the context and available options.  The conscious self enables us to avoid making the same mistake over and over again and it can even enable us to prevent some mistakes before they occur.

Given the remarkable ability of the conscious self to keep us on track, you might ask why we don’t use the conscious self all the time. The answer is simple: conscious attention consumes far more physiological and mental resources than automated action.  Keeping the conscious self in active mode is exhausting. Moreover, excessive use of the conscious self can impede the functionality of the automated self, interfering with our ability to execute even actions that we have previously mastered.

While Kahneman laments our inability to use the conscious self when we really need it, W. Timothy Gallwey sees the relationship between the conscious and automated selves in a different light. In The Inner Game of Tennis, Gallwey praises the amazing capacity of the automated self to learn from experience. Gallwey illustrates his concept of the divided self in the context of tennis. When you attempt to hit a forehand, you enter into one of two modes of action:  a mode of conscious control of action (conscious self) or a mode of natural being and action (automated self). According to Gallwey, our automated self learns and executes with an elegance and efficiency that is not possible for the conscious self. Gallwey maintains that since most of the actions we want to learn are far too complex to be consciously controlled we should entrust the learning of forehands and other complex actions to the automated self. He urges us to trust our body’s natural ability to learn through experience. In Gallwey’s view, it is the meddling conscious self that must be managed because when left to its own devices it impedes the natural learning abilities of the automated self.  Learning is enhanced when we distract the conscious self so that the automated self can learn in peace.

The behavioral economist Kahneman and the tennis philosopher Gallwey both believe that each person is divided into conscious and automated systems or selves. While they both acknowledge the positive contributions of the automated self they are less aligned on the contribution of the conscious self, especially when it comes to learning. While Kahneman sees the conscious self as an auditor that helps us avoid the repetition of stupid errors, Gallwey sees the conscious self as a meddler that must be tamed so that the automated self can work its subtle magic. Depending on the context, each of them is probably correct. Decision-making in tennis moves more quickly than in financial analysis or organizational leadership.  In a tennis match, by the time the conscious self has decided what to do, the ball has already hit the fence and the point has been decided, so there is not much time for conscious control. Still, there must be some role for the conscious self in the learning process of hitting a forehand, shooting a jump shot or having a constructive conversation with a co-worker. Jonathon Haidt suggests a metaphor that clarifies how the conscious self can improve rather than impede the learning process of the automated self.

The elephant and the rider

In The Happiness Hypothesis Haidt suggests that we use the Buddhist metaphor of an elephant and rider to understand the relationship between our automated and conscious selves. The automated self is like an elephant that chooses its course based on a combination of past experience, bio-chemistry and mental associations. In Haidt’s metaphor, the rider sits on top of the elephant and is charged with guiding it toward the right destination. In the real world, an elephant goes pretty much wherever it wants to; the rider cannot force the elephant to go anywhere else. Yet, in the real world, riders do in fact manage to get elephants to alter their course and the way they do it provides a useful analogy for how our conscious selves can alter the behavior of our automated selves. Understanding the relationship between your rider and elephant can open the door to a higher level of personal agency and interpersonal influence.

Automated actions are conditioned actions.

Our automated actions are often triggered by cues that we are exposed to just before the action. When we are exposed to familiar cues, we feel an often unconscious urge to take actions that we have been conditioned to associate with those cues. When I enter a Starbucks, I find it easy to read or write because I have spent thousands of hours reading and writing in Starbucks as a graduate student and mobile professional. In the early years of my junior tennis career, if I saw a ball flying toward my backhand side, my shoulder and elbow would stiffen up because I had already experienced a lot missed backhands by that time.

The rider can’t force the elephant to move, but it can make the elephant feel like moving.

So how do you make your automation abilities work for you? The trick seems to be managing the cues that get your attention before you want to take a certain action. If you can associate the right cues with the right actions, this increases the likelihood that you will be able to consciously introduce those cues at the right time more often. When I see the Starbucks green and smell coffee roasting I feel like sitting down and reading and writing, so sometimes when I really need to get something done I hunker down in a Starbucks. When I listen to Neil Young I feel like reminiscing or brooding on what might have been. This is not an ideal state of mind for focusing on work, so I resist the urge to “keep me searching for a heart of gold” and listen to something else instead. While you cannot control all of the cues you are exposed to in the course of a day, you can control some of them if you choose to do so. The choices your rider makes about what the elephant perceives can have a big impact on what the elephant actually ends up doing in the heat of the moment and thereby on the results you get.

Don’t expect your rider to do the actual walking.

There is limited utility in using the rider’s conscious capacities to analyze and organize every detail of every action you plan to take. You don’t need to consider in great detail exactly how you are going to move your jaw as you chew your next piece of food. When hitting a forehand or shooting a free throw in an actual competition you do not have the capacity to control the mechanics of the shot. Likewise, you can’t consider every single word before you utter it in an actual meeting or presentation. Using your conscious mind to control these actions uses up an enormous amount of physical energy and processing capacity which can actually impede your automated self’s ability to execute the correct actions it has previously installed through experience. In real time, you have to trust that the elephant knows how to do things. The rider’s strength is not execution so much as visualization, preparation and installation.

Think of the rider as the elephant’s coach or visionary leader.

While the rider is not good at managing the details of real-time action, the rider can manage cues to exert influence on the actions the elephant “chooses” to take and how it chooses to execute them.

The rider’s most powerful tools are observation and interpretation. The rider can notice the general patterns of what is going on around the elephant before the elephant does something right. What was the elephant doing?  What was the elephant feeling and thinking about? What external cues were in the environment just before the elephant took the right action?  It is not important or even possible for the rider to record in detail exactly how the elephant did what it did. What the rider should focus on is what was going on that cued the elephant to produce the automated complex behavior leading to the desired result. The elephant knows the mechanics of walking, but it may need help walking in the right direction and cadence.

The rider is the sense maker.

Based on what the rider learns from observing the elephant when it succeeds, the rider can create a picture of success containing elements of what the rider observed in the moments before the elephant did the right thing. The rider can then show that picture to the elephant so that it remembers what things looked and felt like just before it got something right. With repetition, the elephant begins to associate the elements of that picture with the actions that led to success. This creates a cycle of cue-association-reaction that makes the elephant more likely to automatically take the right action in similar situations in the future.

Once the cycle has been reinforced, the rider can create a story or narrative that strengthens the associations deepening the installation of the automated response to similar situations.

Simply put, while the rider cannot walk on its own or hit a serve, it can analyze and organize the conditions that enable the elephant to do these things better. The rider’s influence over the elephant does not come in real time when the elephant is executing. The rider controls the elephant by connecting what the elephant has done in the past to what the rider would like the elephant to do in the future, then creating and re-creating the conditions that cue the elephant to do the right thing. The rider can sometimes interrupt the elephant just before it charges forward and use cues to increase the likelihood that the elephant chooses to go where the rider wants it to go.

There is an old saying that practice makes perfect, but that is not always the case. Practice makes perfect (or at least much better) when the rider learns enough from the elephant’s past experiences to assemble a formula of factors that lead to success then engineers contexts in which the elephant can repeatedly experience the connection between that formula and success.

Between situation and action the rider can automate intention.

There are various mental hacks you can use to intentionally build experience of doing the right thing at the right time. The more repetitions you engineer, the stronger the mental and physical circuitry becomes until the intended action is installed as an automated action cued (sometimes consciously by the rider) in a certain situation.

In subsequent posts I’ll share some of the categories of hacks the rider can use to keep the elephant on course, reducing the frequency of undesirable automated behavior and increasing the likelihood that the elephant’s actions are aligned with the rider’s intentions.


 Thinking Fast and Slow, Daniel Kahneman

 The Inner Game of Tennis, W. Timothy Gallwey

 The Happiness Hypothesis, Jonathan Haidt

The Power of Habit, Charles Duhigg

Talent is Overrated, Geoff Calvin

How to Build a Happier Brain, Julia Beck,  

Need More Self-Control, Try a Simple Ritual, Francesco Gino,

For an exploration of the process of automation (habit formation), see:

In this Ted Radio Hour interview, George Monbiot explains the power of narrative in driving social change in a way that closely parallels the rider and elephant model for driving human behavior:

For a series of interviews on how behavioral nudges can be used to guide the behavior of larger groups of people see:   Freakonomics Radio, How to Save $32 Million in One Hour (Ep. 397)

For illustrations of how pervasive negative framing can be and simple positive narrative reframing techniques see:  TED Talk -Alison Ledgerwood:  “A simple trick to improve positive thinking.”

For illustrations of how operant conditioning provides a model for how your rider can guide your elephant through the process of skill (habit formation), listen to:  “Hidden Brain, Clicker Training for Humans: The Power of Judgment-free Learning”  @

Leave a Reply