Fighting Confirmation Bias Is Like Fighting Gravity (so let’s stop fighting it)

Photo by meriç tuna on Unsplash

When you learn about confirmation bias two things are usually explicitly stated. Confirmation bias is inescapable and that we should do everything we can to escape it. This is like saying it’s impossible to escape gravity but you should do everything you can to try to escape it. 

Is it possible to hack our bias to confirm our beliefs in such a way that we don’t  need to feel like we’re constantly fighting gravity? 

To answer that question we need to understand confirmation bias more fundamentally. 

As I scroll through my social media feeds I feel confirmation bias. More specifically, the posts that confirm my beliefs feel good. They bring me joy. I’m more likely to share them. Those that don’t confirm my beliefs make me feel uneasy, frustrated, or annoyed. I’m more likely to be skeptical of those posts and try to find holes in the logic, which will relieve my discomfort. It will make me feel like I needn’t change my beliefs to embrace this new information. I might even share my skepticism with my followers. 

Confirmation bias operates constantly in background as we go about our lives. If we believe women are bad drivers, we pay attention to examples of women driving poorly and put less weight on men driving poorly. If we want the Tigers to win a baseball game, we will notice when the umpires favor the other team and hurt our team.  If we think “the left” is a cancer on society, we find examples to confirm this and ignore counter examples. Likewise for “the right.” Confirmation bias is always there, telling us where we should direct our attention.

Confirmation bias rides on the rails of the beliefs we have in our mind. In fact, in some sense it’s not so much a bias as it is our brain’s primitive defense mechanism searching for beliefs counter to our own and “protecting” us from them. We perceive ideas that don’t map onto our current beliefs as threatening – hence the negative affect. The negative emotion signals our brain to heighten awareness and seek safety. We find “safety” (hits of dopamine), by poking holes in the ideas attempting to infiltrate our current beliefs. That brings us comfort, but decreases the likelihood we integrate new information that may be true. 

If we are biased to confirm our current beliefs because we instinctively view new ideas as threatening, it’s hypothetically possible to adopt and preserve beliefs that embrace new ideas. That is, we can believe that new ideas enhance our current beliefs. 

Here are a few specific beliefs we could embrace. 

Nobody’s cornered the whole truth on anything

We might start by downloading a belief in our brains that no one person, or group, has a monopoly on truth. This includes you. If we believe this then when we encounter a belief that we disagree with, we will seek out the parts of the belief that enhance our understanding. We’ll seek to integrate both viewpoints, yielding a richer understanding of the given concept. When we find the valid points within the opposing viewpoint, it will confirm our belief (which will literally feel good). 

Essentially, if conformation bias drives us to find information that confirms our belief, then we can set it about the business of finding information that disconfirms our currents beliefs so as to enhance our total understanding what is true. We believe that diverse viewpoints are necessary to a complete understanding, so we seek out those viewpoints, and when we find the truth contained in them, we get the dopamine hit and the reinforcement.  

Our ego isn’t interested in what’s true, only that we’re safe

I’ve read “Don’t Label Me” by Irshad Manji twice and I’m working to become a Moral Courage Mentor. (Moral Courage is Irshad’s approach to psychologically healthy diversity and inclusion training.) In her book and in the course I learned much about the ego, or “egobrain,” as Irshad calls it. This isn’t the “woo woo” ego from Freud – it describes our innate, primitive, threat-detecting system. It’s what’s driving confirmation bias as I’ve described it above. 

Let me pick a totally hypothetical and unrealistic but illustrative example of the ego brain at work (that definitely didn’t happen…). The other day I had a… difference of opinion with my wife. We were getting things ready for a garage sale, which we had discussed would be good to spread over two days, Thursday and Friday. On Wednesday afternoon it became clear to her that we were not going to be ready to have the sale on Thursday. This was not clear to me. As I was cooking dinner she brought up several things we still needed to do to prepare, to which I thought (and probably said) “I guess we’ll stay up late and do them.” She mentioned how she was getting pretty anxious about publicizing the sale without those things done, to which I thought (and probably said), “Well it will be fine, we can get it done.” Finally, she said, “Do we really need to have this sale tomorrow?” 

At this moment I felt my ego say “That’s what we planned and lets just stay up late and do it. Why do we have to change things last minute?!?” However, I’m getting better at noticing when my ego is talking and when my calm, rational mind is talking. This was definitely ego. I thought a bit longer before ejecting the first reaction that came to my mind and realized that it was not necessary to have the sale the next day. My ego wanted me to cling to my old beliefs, my wife presented an idea that was counter to them, and I took the extra second to put my ego to rest, consider the problem more carefully, and ultimately concede that she made a good point. My belief about the importance of sticking to the plan was incorrect and defending it was irrational. 

As I hope you can see, the ego is generating much of our confirmation bias. In fact, we could call our “bias to confirm” our tendency to “protect our current beliefs.” Herein lies another belief we can adopt that would be healthy to confirm: we must routinely speak truth to the power of our ego if we are to update our beliefs about the world. 

Every person is a “plural” 

We have biases towards other people based on the labels they either ascribe to themselves or that we ascribe to them. These labels help us build a caricature in our mind of that person. We reduce them to labels, extrapolate all of their other characteristics from the labels, and then judge them. We quickly categorize a person as someone worth listening to or worth ignoring. 

While this labeling, categorizing, and judging makes navigating our lives easier, it’s unfortunately a house of cards that only fuels confirmation bias. We see a person with a MAGA hat on and we believe we know nearly everything about that person. When he does something that confirms our mental model of a MAGA-hat-wearing person, our belief is confirmed. If he does something that runs counter to it, we either don’t notice or write it off as an anomaly, not to be taken seriously. Pick your favorite tribe to hate on – the same thought process applies. 

However, if we decide to look, we will find that below these labels every person conceals a richer personhood, revealing that they are dynamic and multifaceted. Manji calls a person that consistently bucks their labels a “plural,” and reminds us that if we look (and listen) hard enough we’ll find that every person is a plural. 

If we believe that each person is a plural, then confirming that belief means we pass on snap-judgements and assume there’s more to them than the caricature we’ve built in our mind. 

If we believe that each person is a plural, then we’ll seek the complexity of each person. When we find it, we’ll confirm our bias, thereby reinforcing the assumption that each person is a plural. 

“Wait, you can’t just choose what to believe!”

Sure you can. We do it all the time. Sometimes we don’t realize we’re doing it, but we do. Many times it feels like reasoning leads to concluding that a belief is true, but just as often, if not more, we want to believe something is true and seek out the justification later. 

I think Apple makes better phones than Android makes and better computers than PCs. I’ve got, I think, good reasons to believe this but at no point in my life did I take a year and do an objective analysis on the features of each brand of technology. I had a couple good experiences with Apple products in high school and I’ve been happily feeding that belief ever since. 

More seriously, if we dig into our beliefs deeply enough I think we all get down to a priori assumptions that either consciously or subconsciously adopt. (Books have been written on that topic and I don’t have the space to explore that here. But if you think my argument falls apart because that claim is false, please let me know in the comments!) And, since humans are dynamic, we sometimes change those foundational beliefs. For example, I might go much of my life and assume that people are generally good people. I might then have an experience where I see the dark side of humanity and conclude that people are, in fact, generally bad. 

Is either true in a fundamental sense? 

How would we begin to answer such a question even remotely objectively? 

We can say that adopting either of those beliefs will impact the course of an individual’s life in meaningful ways, right down to daily interactions with other people. I think we can also conclude that a critical mass of individuals adopting either belief will have society-wide ramifications. Finally, in some sense one can choose to adopt either belief – and suffer the consequences. 

Now, not all beliefs are are equally true or pragmatic or will result individual or group-level flourishing. Not all are equally Good. Another book-length exploration would be required to explore what we mean by a Good belief, but for the narrow purpose of this essay let’s assume that it means developing a form of confirmation bias that doesn’t require us to constantly fight ourselves. 

I’m arguing that, given the malleability of our beliefs, we should adopt the the following: 

  1. None of us knows all of the truth which means the other people we interact with must know something important that we don’t – we should listen accordingly.
  2. Our ego often blinds us from the truth in an effort to maintain our current beliefs. This means we need to constantly be mindful of when it’s at work and keep it in check. 
  3. Every person is a plural. This means that we will be slow to put people in boxes and we’ll seek out the characteristics that demonstrate the individuality in their character. 

I’ve adopted these beliefs and I can tell you that my mind is in a healthier place. Remembering that people are plurals keeps me looking for the nuance in their personality. It motivates me to keep looking beyond the caricature I’ve built in my mind. Keeping my ego in check helps me avoid arguments for the sake of being right, as I explained in the story about my wife. I’m not perfect, but more often I find myself taking a breath to respond thoughtfully as opposed to reacting quickly. Remembering that I don’t know everything about anything motivates me to engage with those I disagree with to figure out what I’m missing. Finally, I find myself gravitating towards people who seem to believe the same things. 

In short, in feeding my confirmation bias I gain a richer understanding of nearly everything. 

Thinking, Fast and Slow (Part 3)

Image result for LeeThatcher - "Thinking"

The first two parts of this series were created using Adobe Spark Page. I found while creating this third part that most of what was in it was text, and that a more stable place to host it was probably my blog. Hence, here it is. I’d encourage you to check out Part 1 and Part 2, if only briefly, before reading this part. 

We see in the last few chapters of Part I of the book that System 1 is a story teller. This system helps you make a coherent narrative of the world. This, like most things, has positive and negative aspects. I think the shortcomings of this tendency are important to understand.

The first problem with System one is that it’s relatively easy to manipulate it, especially since so much of it happens automatically. Take the following example from the book. Read the two words below.



In the second or two it took you to read those words and immediately following reading them you had a reaction. Some of it was physical, like the hair on your arms probably stood, your sweat glands were activated, your heart rate went up a bit. But you also likely sketched out a story that involved bananas causing the vomit (or in some other way being connected to the vomit). You did this automatically as system one is attempting to fit the input into a coherent story.

This, like most things, has positive and negative repercussions. It means that we are likely to seek out and find information that fits with the story system one is telling us. “Sally is lazy.” “James is smart.” “Maria is a hard worker.” Once we’ve put these narratives in our mind, system one tries to find information that fits the narrative. And while system two should be the hero here, always evaluating the assumptions of system one, it turns out that system two is a bit lazy. It’s much easier for system two to just go with the narrative. It takes cognitive work to constantly be evaluating everything system one is telling you, so often times that work is avoided by system two.

The key here is that we are aware of the narratives and stories we have in our minds. We need to be on the look out for information that both confirms our narrative (to be sure it does in fact confirm the narrative and that we aren’t overlooking something) and negates the narrative (so that we can change the narrative in our minds to better represent reality).

One major theme of the associative machine is this: when there is some sort of external input to your brain you’re not consciously aware of what’s going on in your brain. When you see an object or hear a sound or experience a feeling, you’re flooded with ideas which in turn activates more ideas. Only a few will pop up in consciousness and this flood of ideas is largely out of your control.

What does this mean for teaching?

When we deal with students we have to remember that fact. Much of the time they (and we, whether we care to admit it or not) are at the mercy of system one. Actions that you take or that other students take in the classroom can set off chain reactions in a student’s brain. The result could be positive or negative. You can imagine starting a lesson with some sort of introduction that primes students for learning, giving them some ideas that ignite other ideas (you might call this engaging prior knowledge) or put them in a positive mood. The latter idea is from Eric Jensen and is also described in the book and that is that being in a good mood helps your brain be prepared to understanding new concepts. You could also imagine taking some action that results in an unwanted behavior. The key to dealing with these behaviors is figuring out what the underlying cause of the behavior is. What is triggering their associative machine to result in the behavior?

The last note I want to make here is that we need to help students understand what’s going on in their brain. Just as I encourage anyone reading this to try to keep a check on the automaticity of system one, we should find ways to help students do this as well. I think doing activities that encourage metacognition is a critical step in that direction, but that’s a topic for another day.

Image is “Thinking” by Lee Thatcher. The original work can be found here.

The Absurdity of One-to-One Initiatives


As comes up every year, someone in our department suggested we go one-to-one. Of course, this sparked lively debate. So much so that do to the frequency of these debates and the cycle of outrage I invariably go through after each one, I’m motivated to write out the multitude of reasons that going one-to-one with textbooks is an absurd idea.

First, let’s talk about costs. A good textbook costs close to $100. Sometimes less, sometimes more, depending on how many you buy. If a kid has five classes, that means it’s going to cost roughly $500 dollars per student to go one-to-one textbooks. And it’s not just $500 one time. Of course not, because in a few years much of what’s inside the books will be dated. They will need to be updated and some of them will be so obsolete they’ll need to be replaced entirely. Do we want to go through the up front costs and then the future costs to update and replace them?

Second, let’s talk about letting teenagers carry around several hundred dollars in textbooks. Have you seen the average teenager’s bedroom? Of course not! There’s too much stuff on the horizontal surfaces (and maybe even the vertical surfaces) to actually see any substantial part of the room. Are we going to let kids, who can hardly get a dirty tissue to the trash can across the room, be responsible for hundreds of dollars worth of school property? I think this is a nightmare that no administrator or teacher wants to deal with.

Third, I’d ask you if the cost is worth the benefit. Sure, textbooks have lot’s of knowledge in them. Give students books corresponding to the subjects they’re learning allows them to easily and quickly look up information, helpful diagrams, maps, and other media, but don’t we have teachers for that? Teachers have much of this knowledge, and if they don’t, then they can look it up in their book and deliver it to the students. What’s the point of a teacher if all the information is easily available to the students? Sure, the best teachers might use the books in coordination with their teaching ability to create near sublime learning experiences, but this would surely only be the most motivated of teachers, not the norm.

Fourth, it would be a logistical nightmare. Suppose your high school has 1000 students. Then we are talking about THOUSANDS of textbooks to keep track of. Not just keep track of, but record those that go missing and those that are damaged. Then schools have to make determinations about how much the damages cost. Then who pays for it? The students? What if it’s an accident? What if the student can’t afford it? What if they get lost in a house fire? Who’s on the hook for the bill then? And who does this burden of tracking fall upon? The library? The administrators? The teachers? There’s no good options. And then, who fixes them? Do we offload this responsibility on the already short staffed library personnel? We’d probably have to hire somebody to spend part of their day repairing textbooks, so tack that on to the bottom line.

It seems clear that the costs of trying to put the these resources into the hands of each student almost certainly outweigh the benefits. But this fight will never die. Year after year we’ll continue to hear how we should “give students access to the tools they’ll be using when they leave us”. Given what I’ve outlined above, I can’t see the logic that results in this being a good idea.

Update: I should make something clear. This is purely satire. I am simply trying to make the argument that when it comes to discussions of 1 to 1 technology I think the problems that are brought up are often ones that we have solved in other contexts. This situation never came up in my department. And even if it had, I would never throw them under the bus like this publicly. Once again, this is purely satire.

We should always be learning something new

Last week I started auditing a class on Hapkido, which is a Korean martial art. A friend asked if I’d be interested in attending the once-a-week class with him and I said I would. Having never done any martial art at any time in my life (I don’t even think I’ve watched any of the Karate Kid movies in their entirety), I was nervous. However, partway through the class I realized that there is clear value in learning something completely new.

I want to list a few feelings I had, as I think they reminded me of what it’s like to be a learner, as an adult or a teenager.

  • I didn’t want to make any mistakes. When the instructor demonstrated something, I wanted to do it perfectly. This notion is ridiculous because, as the instructor also pointed out, it takes thousands of repetitions before something becomes muscle memory. For as much as I preach the importance of mistakes in learning, I was shocked at how somewhere in my guts I still didn’t want to make them.
  • I didn’t want the instructor to come by me. Or at least if he did I wanted to be working on my right side (which I thought I was better at). I was afraid he’d find something I was doing wrong. Which I consciously knew would not be bad as it would get corrected and then I’d improve.
  • I compared myself to the people around me, unconsciously ranking myself. Better than that person, worse than those two, etc.
  • Frustration. I’ve never been particularly coordinated and I was consistently frustrated at knowing in my mind what I wanted my body to do, but struggling to make my body do it.

I walked off the mat at the end of class and my mind was reeling.

“Now I know why students are apprehensive to ask questions.”

“Now I understand better why a student might get uncomfortable while I hover over them watching them work out a problem.”

“I have to constantly remind myself to embrace the difficulty. That’s where growth comes from, but it’s difficult to do in practice.”

“Having an instructor that recognized we were all learning was incredibly helpful. He created an atmosphere where mistakes were not viewed as setbacks, but part of the process.”

My main takeaway was that these are feelings I need to constantly grapple with. I need to try to put myself in situations in which I’m the learner, with relative frequency. It helps me better understand where my students are coming from and I think will ultimately help me become a better teacher.

Also, here’s one more thought that has popped in my head recently and probably doesn’t need an entire blog post, but fits with the theme in this one. I’m in my fifth year teaching precalculus, AP calculus, and algebra II and I can feel myself having less empathy with my students, with people learning the concepts for the first time. The first year I taught these courses I think I had a better understanding of their struggles as I was solidifying my understanding of the concepts prior to teaching them as well. I’m not entirely sure what this means for my teaching now, but I think awareness of it is important.

“Everything Springs from That”

I don’t listen to many political podcasts. In fact, only one. Dan Carlin’s show, Common Sense. In his latest podcast he interviews James Burke, a science historian, documentary creator, broadcaster and all around smart dude.

This episode flirted with politics, but was more focused on how technology affects society and how the rate of change often has unforeseen ripples. It’s a fascinating interview, but the best part for me comes at the end of the interview. Dan presents Mr. Burke with a hypothetical (which I’m paraphrasing).

Suppose the leaders of the country call you up and ask for your advice. What would you tell them in regards to the absolute most important thing to focus on in the future?

“I’d say put a massive amount of effort into the educational system. Everything springs from giving people the kind of education that allows them to think more clearly and express themselves more clearly. Everything springs from that.”

I’ve been thinking about education a lot lately. I recognize that might be like pointing out that a historian has been thinking about history a lot lately. But I’m talking about the big picture of how we educate our society. With the appointment of charter school evangelist Betsy Devos to the head of the Department of Education and recent moves by the Michigan congress to weaken the teaching profession and cut funding, I worry greatly about where we are headed.

The election of Donald Trump, the proliferation of fake news, the gravitation towards soundbites, the lack of empathy, and constant decrease in social capital mean that having a society that can’t think critically could be (already is?) disastrous. If there was any time in our history that we should be focused on education, it should be now.

We can’t have a society of mindless drones that will believe the headline and first two lines of any article that comes across their news feed. We can’t have a society that can’t take another person’s perspective. We can’t have a society that fears change. We can’t have a society that doesn’t understand the value of civil discourse.

An education system that’s working on all cylinders can help prevent this.

We should be focused on how to graduate great teachers. We should be focused on how to help teachers become great. We should be looking to other education models and schools that we want to emulate. We should be focused on making teaching a profession that our best and brightest want to pursue. We should be working to get away from standardized test scores as the sole measurement of a quality education.

As Mr. Burke mentions in the podcast, if we put as much energy and money into education as we did into the Apollo project it could have countless dividends for our society.

Order – How Mathematics is Life

Humans are in a constant pursuit of order. We try to develop schemas to help us deal with frequently occurring situations. We constantly look for patterns. We try to make our lives somewhat predictable.

The brain doesn’t like to think. Thinking is hard. So the brain naturally gravitates towards pattern finding.

This is mathematics.

Mathematicians look around the world for patterns. Looking for truth. They take things they know to be true, and build on them. Constantly growing the body of patterns we know to be true.

The difference between me noticing that whenever it’s cloudy out I’m a bit gloomy and that the derivative of a parabolic function is linear, is that the latter is true always. It’s a fact that exists regardless belief, mood, perspective, or measurement.

I wrote the idea for this post down months ago, but it seemed relevant as this week I embarked on teaching my algebra II class how to factor polynomials. Something that nobody does, with the exception of math teachers and their students. (And I mean that quite literally. I went to the twittersphere and came up empty.) My advice to students was similar to other seemingly obscure content we learn in mathematics.

Treat these problems like puzzles and look for the patterns.

Because pattern finding, curiosity, and creativity in problem solving are all skills that are valuable and can be improved with practice.

Nobody does a puzzle and while they’re doing it says, “This is never going to help me in my life.” I don’t claim to be an expert on the motivation of puzzlers, but I did puzzles just to figure them out. I enjoyed the mental exercise.

This is how I want my students to approach math problems. I want them to enjoy and appreciate the pursuit of solving the problem. I know that’s abstract and might be difficult for teenagers to grab onto, but I’m not sure of any other justification for some of the concepts we teach.


Thoughts on “GPSing our Students”

In June Dan Meyer posted Your GPS is Making You Dumber and What that Means for Teaching. In it he makes the argument that providing step by step instructions for math concepts results in students being able to get from point A to B, while not understanding much about the concepts they’re supposed to be learning. His argument can be summed up with this paragraph, and is somewhat inspired by what Ann Shannon wrote in what teachers should Look for in the CCSS Mathematics Classroom.

Similarly, our step-by-step instructions do an excellent job transporting students efficiently from a question to its answer, but a poor job helping them acquire the domain knowledge to understand the deep structure in a problem set and adapt old methods to new questions.

I would tend to agree. I do give students steps occasionally but it’s often in order to simplify concepts and, if I’m being honest, to some degree avoid students truly struggling and grappling with the concepts.

I’m curious as to what others think about his post and the notion that GPSing students leads to less learning.

Panera Bread and Learning

In the last month I’ve become a Panera regular. We’ve been doing a lot of traveling and Panera was as close to fast food as we were willing to go. Until the last month however I’d only been there a few times in several years, without being particularly impressed.


image credit:

Take a look at the menu above. This is what you encounter when you walk into Panera Bread for the first time. Seven sections of menu packed with different dining options. For the less food savvy among us, not only is the menu packed with stuff, but it’s packed with a lot of stuff that is unfamiliar. In my first couple of visits I stood there bewildered for a while, let my friend order, then picked something with turkey in it.

Because I was pretty sure I knew what turkey was.

Well, I was wrong. I mean, I think the thing had turkey in it. But everything else did not complement the turkey in a positive way.

The next time I went there with a few colleagues I ordered a caesar salad. Why? Because every time I’ve ever ordered a caesar salad in a restaurant I haven’t been surprised by what came out. The same was true for this visit.

Sweet. I found something on the menu without holding up the four people behind me.But after that I avoided Panera for a while. A long while. I could get a caesar salad from a lot of places. The notion of standing there in line trying to find something different felt as though it would only bring frustration.

Then my wife and I were looking for something quick and healthy to eat, so we decided, to my slight dismay, to give it another try.

This time however, on the way there my wife looked up the menu on her phone and read off several items, describing them from their website. When we walked in I knew several of the items on the menu, how it worked (that “pick 2” thing for example) and was basically ready to order when I walked up. We’ve been there a couple of times since and now it’s one of my favorite restaurants. Since I’m confident in my understanding of the menu, I try new items, ask questions about different types of food, etc.

How does this relate to learning?

Think about the first time you encountered an unfamiliar topic and tried to learn it. This was difficult at first but I can certainly think back to math lectures (here’s lookin’ at you, Linear Algebra) in which there was new notation, vocabulary, and concepts and it all felt unfamiliar.

However, once I got with my peers and we began working through the problems (encountering the “menu” multiple times) the concepts began to feel more familiar. I realized that I knew more about them than what I thought (romaine and kale! Hey, I know what those are…). The more I worked individually and with my peers the clearer the concepts became.

If I avoided doing the exercises or only did it individually, that feeling of everything being foreign never really went away. Usually parts of the notation would be confusing. Or the instructions around a problem wouldn’t make sense. Or I could start a problem, but get lost in trying to solve it, etc.

A couple points can be pulled from this. First, a student’s first exposure to a topic is incredibly important. If you drop seven new vocab words and a gaggle of new notation on students at 8:00am on Monday morning you’re bound have a large group of students not wanting to come back to the menu you just presented them.

Even if they should know most of the food on there.

So we have to think carefully about how students first engage with content. The second point is helping students understand that multiple engagements with a concept will (usually) alleviate this feeling. I think many students never go back to the restaurant because they don’t want to be embarrassed for not knowing what their peers may already know. We have to help students be comfortable with this phase of unfamiliarity (my study group in college was a place I felt comfortable being wrong), and develop tasks that help them engage with the concept in safe, productive ways (investigating the menu on the way).

For a great post on how students see unfamiliar mathematics check out Ben Orlin’s recent post entitled What Students See When They Look at Algebra.

Alternative Titles

“What’s on that sandwich? Oh… will that be on the test?”

“How Avocado Ruins Education”

“I don’t even like lettuce so why should I be learning about salads?”

Educating in a world of Robots

The problem with the way we educate and the things we value in education is that it isn’t preparing students to work in a world of robots. I don’t mean C3PO and R2D2. I’m talking specifically about a world in which most people have access to and actively utilize artificial intelligence.

What’s artificial intelligence?

This is important to understand before moving forward. When I talk about artificial intelligence (AI) I’m talking about machine learning. The idea is that a computer (or several computers) can take a large amount of data, run algorithms and software on that data, and use it to make decisions and predictions. AI already exists in many of our lives.

  • Have you ever asked SIRI, Google Now, Cortana, or Amazon Echo anything? – AI

  • You know how Google photos (and other software) looks for faces in photos and groups them together accordingly? – AI

  • Remember when that IBM computer, Watson, beat those guys in Jeopardy? You guessed it – AI

  • Cars that drive themselves? – AI

Here’s a few other ways AI probably also impacts our life right now.

Before I get into how this should change the way we educate, you need to listen to this four minute segment of The Tim Ferris Show podcast in which Tim interviews Kevin Kelly, the co-creator of Wired Magazine.

Go directly to the part of the interview where Kevin talks about how artificial intelligence will be as disruptive as the industrial revolution (by clicking the link), come back to this post, and read on.

If what Kevin says is right, and it probably is, then we need to drastically rethink how we educate and what we value in education. Here’s a few skills we need everyone in society to have in a world that’s drastically different from anything we’ve ever known.


Answers are easy and will only get easier.

You may be thinking, “not all answers are easy.”

You’re right. But many of the answers that are difficult to answer are generated by quality questions first. These questions are often deeper than most realize. In fact, the question is often the most important part of the process. If you’re answering a worthwhile question then it’s likely that you’ve spent some time framing and developing it. Take this example Warren Berger cites in his book A More Beautiful Question.

The developing world has a shortage of incubators. For years, health organizations and philanthropic groups asked the logical question: How can we get more incubators to the places that need them? A relatively straight-forward answer to that question was – donate them. But that was the right answer to the wrong question. This led to thousands of incubators being donated to poor nations, “only to end up in ‘incubator graveyards,'” as the New York Times reported. … The better question, which was eventually asked by health officials working on the problem, was Why aren’t people in the developing world using the incubators they have?

As it turned out, the problem was that people in the developing world didn’t have parts to fix the incubators (and other donated medical equipment) when they broke. The solution became to build incubators out of mostly car parts, as these were much more abundant.

Had the question not been reframed people would’ve continued with the easy answer to the wrong question. The technologies that will develop over the next decade will require a society that can ask the right questions.

And if you still don’t believe me, here’s an Einstein quote, and here are a few more if you still don’t buy it.

“If I had an hour to solve a problem and my life depended on the solution I would spend 55 minutes determining the proper question to ask for once I know the proper question, I could solve the problem in less than five minutes?” – Albert Einstein


Consider what computers can do now. They can drive cars, analyze photographs, win chess games, and predict the song you want to listen to next, among other things.

Consider what computers can’t do now. They can’t design a more fuel efficient car. They can’t take a photograph of your child that’s worth hanging in your living room. They can’t develop new games. They can’t write a song.

They can’t create. At least not yet.

This leaves the creation to people. Everybody says this and it’s because they’re right and it’s true.

Assembly line work is dead or dying.

We can’t educate for that world. We need to educate for a world that we don’t even understand and can’t predict. But given the information we have now it certainly seems that people that can create, or work on teams that create, will be the ones that are least likely to be replaced by robots.


Technology will bring with it many solutions but also many problems. More and more society will be at the mercy of algorithms.

Consider how Facebook tweaked news feeds to determine how it affected people’s moods. Are we, as a society, okay with that?

Kevin Kelly brought up an interesting point in regards to self-driving cars.

Should the car favor the driver or the passenger in an imminent accident?

How far do we take our knowledge of genetics and human genome modification?

To what extent do we sacrifice our privacy in the name of safety?

What are the ethics of making physical attacks through computers?

There are a plethora of other questions that will arise. With great change comes difficult questions that society will have to answer. Members of society will need the ability to reason logically. They need to be able to reason in contexts where there is no simple answer. Students that think most problems are simple and straightforward often become adults that think the same, failing to realize the complexities in problems. A healthy democracy thrives on a society’s ability to have healthy, rational discourse. Anybody that’s seen social media or cable news can see this slipping away quickly, and I fear a society that can’t think beyond the soundbites will fall to those that know how to control them.

We can teach these skills in classrooms. We can develop our classrooms in ways that foster questioning, creation, and reasoning. But this means that many of us need to shift how we teach what we teach. We can’t simply continue to value only answers, giving students the illusion that the world they’ll live in will be simple. A world in which the important questions don’t have four choices and a bubble to fill in.

I’m not arguing that knowing is dead. But “knowing stuff” will certainly become less and less valuable as knowing what to do with the stuff you know becomes more valuable. The skills that we need to know (questioning, creating, learning, reasoning, etc.) help us connect the knowledge “dots” in the world. It will be more important to know how to attain certain knowledge than to have/store it. A person needs to be able to find information, skeptically analyze it, then integrate it and apply it to the information they already have.

If we can’t help students develop these skills then we’ll fail as educators. Our job is to prepare our students for the world they’ll live most of their lives in. In many ways that world is shrouded in unknowns. But the skills I’ve outlined above will almost certainly always be needed to thrive.

I am not going to be the one that buries my head in the sand and in 20 years says, “I didn’t know it was coming.”

And I get that t’s scary. Assembly line education makes sense and is straightforward. It’s what most of grew up with. It makes sense.

Teach vocab word. Practice vocab word. Test vocab word. Know vocab word. Learn next vocab word. And so on.

See nail. Grab hammer. Strike nail.

The problem is that computers are good at that. Better than people and only going to get better.

The scary part is that teaching the skills outlined above is fuzzier. It’s difficult to isolate it down to a data point. It requires teachers with knowledge of how to teach the skills and trust in the teachers to teach them. This is scary for teachers and administrators and probably most people in education. But it’s the skill set that the average member of society will need in the near future and it’s our prerogative to teach it.

Experience != Improvement


I’m five years into this teaching gig. I’ve taught the courses I’m going to teach next year for the last four years. I have lesson plans for each day, multiple forms of assessments written, and only a few notes scattered around for things to tweak next year. I’d like to explore the problems with being at this point in my career, specifically that it would be easy for me to practice teaching, without getting better at teaching.

The notion that practice by itself doesn’t make you better dawned on me last summer when I was taking classes towards my master’s in edtech. We read most of a book called “Why don’t students like school” by Daniel Willingham. In chapter nine, entitled “What about my Mind?”, the author draws the distinction between experience and intentional practice. He notes, “Experience means you are simply engaged in the activity. Practice means you are trying to improve your performance” (p. 92). This means that simply because you have years of experience doing something doesn’t mean you’ve been improving.

Willingham uses driving as an example, saying, “Like people my age I’m experienced – that is, I’ve done a lot of driving – but I’m not well practiced, because for almost all of that thirty years I didn’t try to improve” (p. 92). I’ve noticed this in my own practice and I’m only five years in. I’ve taught many classes the second or third time and thought, “that didn’t seem to go much better than last year.”

Willingham argues that there are three facets to improving as an educator.

  1. Seek out opportunities to improve your skills
  2. Consciously try to improve your teaching
  3. Gather informative feedback


I find that I’m pretty good at the first one, okay on the second one, and not good on the third.

To be clear, I’m not saying that you should be pouring hours into improving every lesson of every day all year or that everything should constantly be in a state of flux. That would be exhausting. What it does mean is that you can’t just repeat the same things over and over expecting improvement to just happen.

Deliberate Practice

You may have heard of the 10,000 hour rule as made popular in Malcolm Gladwell’s book Outliers. The idea is that to become an expert in anything you need to practice it for roughly ten thousand hours. What sometimes gets lost behind the number is that it has to be deliberate practice, which is a very specific kind of practice (described in detail below). Gladwell describes this kind of practice but Dr. K Anders Ericcson is the person that did the initial research on it. If we aren’t engaging in deliberate practice, especially once we’ve gained a basic proficiency in a skill, we are likely to stop improving. In this I think Willingham and Gladwell and Ericcson are in agreement.

Here’s an example that demonstrates this. I’ve played my guitar a lot in the last couple of years but have not improved much. Why is that, despite the fact that I play at least 15 minutes every day?

The reason is that, as my wife kindly puts it, I’m mainly just “plunking around”. In other words I’m not working on specific techniques or skills each day. I’m merely playing things I already know how to play and am not actively stretching my abilities. I reached an ability level where I can play most of the stuff I enjoy playing, and I’ve coasted. In fact, I’ve slid backwards in a lot of ways as I think back on songs or riffs that I’ve forgotten.

Corbett Barr, creator of Expert Enough, in his piece Deliberate Practice: What It Is and Why You Need It, provides a good description of deliberate practice. He defines it as “a highly structured activity engaged in with the specific goal of improving performance.” This requires identification of weaknesses followed by engagement in activities to improve on those weaknesses. This is clear contradiction to my guitar practicing, and Willingham would argue in clear contradiction to how many veteran teachers approach their profession.

In the first few years of teaching, improvement is paramount. Teachers make lot’s of mistakes early so the amount of time spent reflecting and working to improve is plenty. Like any learning curve, it’s steep early on by necessity. After a few years a teacher is likely to have the big things (management, curriculum, assessments, etc.) generally worked out. It becomes easy to fall in the trap of just “plunking around”.

And this is not teacher shaming and I’m not claiming that more experienced teachers are worse than younger teachers. I’m not claiming all teachers fall into this category of not engaging in deliberate practice. I recognize that we are all generally trying to get better, but we have to be aware of how we are going about this. It’s easy once you’ve found a system that “works” to assume that the more you do it the better you’ll get.

If you’d like to learn more about deliberate practice I’d encourage you to listen to the The Freakonomics podcast episode How to Become Great at Just About Anything. It contains in-depth discussion about how to improve at anything. Much of the discussion in this episode also runs in the same vein as the concept of growth mindset. You can listen to the episode right in your browser with the above link, or head over to the Freakonomics page for the audio file and show notes.