Learning is a rather straightforward process that involves six 1st principles; awareness, context, application, feedback, repetition, and iteration (read this article to learn more). And while all these 1st principles are critically important to learning, the real magic happens when we go to apply the knowledge that we are attempting to learn. The application is what makes learning exciting and what creates lasting memories of what we are learning. The application is when we go from a purely static concept to something dynamic that brings us closer to the truth.

The reason that our understanding of the truth must be dynamic is because truth itself is, has been, and always will be more complex than we can understand (read this article to learn more). While we can get closer to the truth we will never fully arrive. We know that atoms are mostly empty space and yet we can’t just willy nilly pass through them. There’s something solid there even if we don’t fully understand how it works. By rendering our understanding of the truth into something dynamic it allows our understanding to become a process rather than a static outcome. This allows us to move closer to the truth over time. That same dynamic nature of the application is what allows us to correct any mistakes as well. None of this is possible when our understanding of the truth is merely the static outcome of rote memorization.

The application is when a concept is inputted into a system of one kind or another to see if it works. Whether that’s telling a joke for the first time to see if it gets a laugh, or trying to apply a new method for solving a math equation. These are all applications to see if our understanding of the knowledge that we are learning is correct or not. The application is the moment of insight, whether that be an abstract application or a concrete one, when we apply a piece of knowledge, we are getting closer to its relationship to the truth. A correct application means we got closer to the truth and a wrong application means we learned either a piece of knowledge or application of that knowledge that doesn’t work.

In this way, applications allow for genuine feedback that can lead us closer to the truth over time. We can parrot what a book tells us about riding a bike but, we have no way of verifying if it’s right or not simply by reading it. This changes when we try riding the bike ourselves. Whatever mostly static concept of riding a bike was in our mind’s eye gets immediately converted into a dynamic mental model when we start applying said knowledge. Once we start riding, we will get concrete feedback about our understanding, if it works, and a flood of nuances that make for a better mental model. That same dynamism also enables our mind to connect the new application to other knowledge and skills via overlapping mental models.

This is why teaching anyone a giant compendium of facts and figures about economics, history, government, and so on doesn’t work. Our understanding of history is constantly changing and adapting as we gather new evidence and discover new forms of reasoning we can apply. But when we learn these subjects as static entities, we don’t know how to adapt our thinking and understanding when new evidence comes down the pipeline. We also don’t know how to connect it in any meaningful way. Instead, we desperately cling to whatever view we have trying to stuff facts and figures into a malformed and static model. This leads to dogmatic and ideological thinking that must reject new findings, innovations, facts, and reasoning skills if they disagree even slightly with whatever bias we are operating from. This is what a static understanding is. A model of learning that quickly reaches a point where dynamic learning is no longer possible. With static models of learning, nuance all but becomes verboten.

Equally important to applications and the feedback they can generate is found in the power to highlight our mistakes. Many people talk about learning from failure, but this is wrong. Failure is what happens when we fail to learn from our mistakes. Mistakes are a natural way to focus our mind on some part of the application or knowledge that we need to increase our understanding of. Put another way, mistakes are where we need better learning. Here again, the application reigns supreme. It provides a vehicle to highlight our mistakes and allow for possible solutions thanks to its dynamic nature. And once a solution is found, we can practice, aka repeat that solution until we are certain that we can apply it reliably. We repeat it until we are certain we have learned it. And from there we iterate. In short, the application that provides feedback allows us to find and correct our mistakes in this process I call dynamic learning.

Dynamic learning is the process of using an application and feedback to find and correct our understanding. The trick to learning from our mistakes is to understand the difference between productive mistakes and unproductive mistakes. A productive mistake is a mistake we learn from. An unproductive mistake is a mistake we don’t learn from. The mistakes that we don’t learn from despite having the opportunity to do so shows us where dogmatic thinking has likely taken hold. Everyone, myself included, suffers from this. We all have the same mistakes we’ve been making for years on end with little to no effort to correct them. Quite often, we are even hard pressed to admit we are making them. While dogmatic thinking is likely the majority culprit, it is sometimes the case that we simply don’t want to invest the time and work to learn a better way as well.

Then there’s the unproductive mistakes we can’t learn from. These mistakes either knock out our ability to learn or exceed our capacity to learn from. If we get in an accident and kill ourselves, get fired from a job before we could properly learn a skill, etc then our mistake knocks out our ability to learn. Obviously, these mistakes should be avoided at nearly all costs. Then there are the mistakes that are too difficult to reverse engineer and learn from. If we try to make highly accurate weather predictions a month out, the mistakes we make are too difficult to reverse engineer a solution from. In such an event, we are well aware of our mistakes. The problem is that, correcting, aka learning from such a mistake is literally too hard to correct. Even if we want to learn we literally don’t know how.

Unproductive mistakes can also happen when we make too many otherwise productive mistakes in rapid succession. It’s not that hard to throw a single punch with good form. But to throw a large combination of eight punches in rapid succession is something very different. We can easily correct for a single punch that’s been thrown with poor form. But correcting a giant flurry of mistakes as every one of the eight different strikes is riddled with technique errors becomes nonsensical. The mind won’t even remember what it did on any one of them, let alone all eight. Nor will a good instructor be able to recall all the problems. If the novice student had only worked on one punch, the mistakes would’ve been productive because the number of mistakes would’ve been manageable.

This is why learning depends so heavily on what I call a productive struggle. A productive struggle has many different names such as desirable difficulty, zone of proximal development, and Goldilocks zone to name but three. But all of them amount to the same thing. A productive struggle is a relevant difficulty that creates more productive learning. As we will see, this works best when housed within an application. When an application gets too easy, we are likely not learning very much. And if it gets too hard then we start making unproductive mistakes which also means we aren’t learning very much. While making a steady but not unmanageable stream of productive mistakes is probably the best evidence we have that we are creating a maximally productive struggle, any increase in difficulty beyond the very easy is enough to begin increasing our learning curve.

Productive difficulties are so important to learning that increasing the difficulty of almost anything that seems easy, particularly that which was otherwise really easy will almost universally lead to a better outcome. This is why I chose to write this paragraph in italics. It makes reading this paragraph slightly more difficult. The slight increase in difficulty means you will better learn it. Even something as simple as using a more challenging font is enough of a difficulty to produce a boost to our learning. This is one of the rare things that virtually everyone in the field of learning agrees with despite the ubiquity with which it’s ignored in education. If we make our goal to create the easiest pathway to learning which is increasingly the case, we are actually ensuring that almost nothing will be learned. This is Likely the bIggest danger with AI being introduced in education. The most effective learning always comes from the most productive struggle, not the easiest application.

We can even see the explicit use of a productive struggle in more advanced forms of repetition. Spaced practice is a form of practice that intentionally introduces greater gaps of time between practice sessions such that we begin to forget a little bit of what we were practicing. It consistently outperforms a more standard practice model that emphasizes greater frequency. It necessitates an added focus on what we are learning that produces better results both in terms of more efficient use of time and in absolute terms of better learning (read this article to learn more).

Interleaving is another method for introducing a productive struggle into the application. Interleaving is when we constantly switch between two or more different iterations of the same skill or similar skill sets. By constantly switching between different iterations that are still highly related to one another via their skill set, we force our mind to better calibrate each iteration. This causes a productive struggle by a different means than spaced practice. Free throws give us a really good example of how this works.

Imagine that you wanted to get better at free throw shooting. The standard model would be to stand at the free throw line and take shot after shot slowly getting better over time through very small technical changes and tons of repetition. This is known as blocked practice and is the natural way in which people seem to approach practicing almost everything from golf shots and free throw shooting to math and more. This feels intuitively correct to us because our brain does this naturally when it organizes information and skills into relevant blocks known as chunks. Blocked practice is a way of isolating whatever chunk we were trying to learn in order to focus on it. The problem this creates is that progress slows to a crawl because of how quickly there is little to no productive struggle. In other words the repetition becomes rather mindless.

When we interleave free throws, we don’t just take them all from the line. We interleave between the free throw line, one foot in front, and one foot in back. By constantly changing where we take each shot, we force our mind and body to adjust. Making so many adjustments so quickly amounts to creating a productive struggle in the form of recalibration. Even if it doesn’t create a lot of mistakes, it still makes it harder to shoot a free throw. That increase in difficulty is what pushes our application of free throw shooting to a higher level faster with less repetition. Blocked practice can do this but only if we are just beginning to learn how to shoot a basketball. In the early stages of learning any skill, blocked practice will typically create a productive struggle, due to the novelty of what’s being learned.

Interleaving isn’t just relegated to physical skills either. Studies have shown that this works really well for things like mathematics also. A student will have to work out different types of math problems on the same worksheet. One will be a long division problem while the next might be the multiplication of fractions and so on. Whether a physical skill or an abstract skill, interleaving increases the productive struggle by creating a constant need to recalibrate the application we are learning. Notice that one cannot recalibrate memorization work by very much. It’s rote recall, which means it’s static. Sure, it will help a little if we break up the order in which we recall facts and figures, which is why people regularly shuffle the index cards they use as a study guide, but this isn’t interleaving as it doesn’t create a need for recalibration.

But, there is one catch about interleaving the application. If we make the various applications too different, like trying to mix long division and free throw shooting, there’s no benefit. In fact, such attempts at learning actually deteriorate our learning curve, rather than improve it. Keeping the iterations we are applying similar allows our mind to stay focused on the same basic circuitry so to speak. We aren’t trying to completely change gears. In this way, interleaving avoids the mistake made by multitasking. It keeps it in the same general vein but forces a greater difficulty that improves our learning curve via the application. Switching between math and free throws is just multitasking, which universally harms learning no matter what form it takes.

By introducing the concept we wish to learn into an application we make what we learn dynamic. That dynamism is what allows our mind to connect it, accurately, to other concepts and skills as well as gather valuable feedback. That feedback is the vehicle to find and learn from any mistakes that naturally happen during the formation of our mental model. The dynamism that this creates along the way is how we get closer to the truth over time.

The application also provides us with the perfect modulation tool to create a productive struggle. Using tools like interleaving works far better than spaced practice because the difficulty we add to our learning doesn’t depend upon forgetting what we have learned. In this way, spaced practice requires that we stop learning to be effective while interleaving gives us the ability to learn so long as the will and energy to do so are sufficient. This allows us to engineer a struggle that will improve our learning.

If we take away the 1st principle of the application and our ability to modulate it to produce a productive struggle, learning falls off a cliff. Sadly, this increasingly appears to be the norm in education today. We are increasingly taught to create overly simplistic and static models of the world. These models are destined to become the dogmas and ideologies, even if they weren’t designed with either one in mind. Tragically, it’s becoming clearer each day that dogma and ideological thinking are intentionally added to what’s taught in K-12 and higher education today. That lack of both a dynamic nature to our learning and the productive struggle that can be built therein ensures that we will get further from the truth over time. This is literally the opposite of what the learning process is so clearly designed to do. Learning is humanity’s superpower, but only if we use it to get us closer to the truth. Anything else will likely only hasten our destruction.