Knowledge is good; skills are better. Why do we have to understand everything?
Do you think the saying “Once you learn, you never forget” is true? Only partially. Think of the countless Spanish or Latin vocabulary you learned in school. How many of them do you still understand today?
On the other hand, think about cycling. Even if you haven’t ridden a bike in years, you can always get back on and go. Why is it that we remember some things forever and others only for a short time?
Here’s the thing: Once we learn something, we can unlearn it. But once understood, we can’t un-understand it.
In school, vocabulary lists and rules are preached: “And now you’re going to learn this by heart.” Cramming is incredibly dull, demotivating, and contradicts our natural urge to discover. It also objects to the automatic learning processes in our brain.
We don’t understand by memorizing something. Nor do we understand by continually repeating the same thing. We need to grasp the underlying meaning to learn sustainably.
We can accumulate knowledge, but we will forget it after a while if we do not actively use it. Only what we grasp makes sense to our brain and remains stored for eternity. And only what we use again and again becomes an “ingrained” skill.
If you don’t just want to learn the world by heart but also understand it, you shouldn’t stoop to the level of computer algorithms.
Artificial intelligence, deep learning, and what they are all called: These terms make us marvel these days because they promise the highest intelligence. Are these techniques even more intelligent than we humans? Can we learn from them so that we can acquire skills more quickly? No. Because these algorithms are continually calculating away, without sense or reason. That is a big and essential difference to us humans.
Machines are constantly calculating away. On the other hand, we humans take a short mental break when we have seen or heard something new – and activate those brain regions that grasp the meaning of further information. Our brain classifies the new information into meaningful or meaningless. Only information that can be linked to existing data in the brain is classified as useful. This data is integrated into the neuronal network. Neurons fire, connections between countless neurons are established, expanded, or strengthened. In this way, a kind of road network is created in the brain. The more often a road is traveled, the wider it becomes – at some point, it may become a highway.
So during the breaks, our brain classifies and processes the information. Most of the time, we don’t notice anything about these brain processes because they happen entirely automatically. These moments of rest are necessary for understanding. Does this not seem like a nice resolution?
The use of prior knowledge remains the greatest advantage of humans. We are able to understand complex concepts with quite little data.
Deep Learning teaches machines to learn. It can thus constantly improve its capabilities independently and without human intervention. The model is the human neural network. Based on existing information, the system can repeatedly link what it has learned with new content and thus learn more. Think of voice assistants like Siri and/or Alexa. These systems can independently expand their vocabulary with new words and phrases.
Thanks to Deep Learning, even very complex issues can be recognized and assigned, and problems can be solved. Artificial intelligence can already solve some subtasks better than we can today. But Deep Learning does not offer unlimited potential. The problem is obvious: the software requires enormously large amounts of data.
Reality is incredibly complex and diverse. Humans are very good at processing information in context. As a result, we need far fewer amounts of data to recognize a connection or solve a problem. Intuition also plays a role here. We move in this complexity every day, and – unlike software – we are not in a closed system.
If we see a red light in a neighbor’s garden at night, we recognize it as party lighting and do not wait for the “traffic light” to change to green. A toddler learns what a duck looks like in a children’s book and recognizes it very reliably in real life, as a stuffed animal, or on television – even without an explicit rule definition, and it has no problem with a slight variation. This is not the case with deep learning algorithms. This is because they can only learn such correlations based on a very large number of examples. If these large amounts of data are not available, then these algorithms have a very hard time. The quality of the prediction results drops.
The use of prior knowledge is also necessary to distinguish logically meaningful correlations from random correlations. An example: The birth rate in Western European countries is steadily declining. At the same time, the number of sighted storks is declining. Are there now fewer babies because there are fewer storks? This is, of course, an obvious fallacy. Just because an apparent correlation exists does not mean that the data are logically related.
3 Learnings and study tips for sustainable understanding
1. Learn in Bites
Think about what you want to learn and divide the material into bites. It is better to study in many small steps than a few big ones. Our brain continues to learn for about 7 minutes after each intake phase, even – or precisely because – we have been concentrating on other things for some time. We don’t realize it, but our brain gives us 7 minutes of learning period each time. So divide your learning into 10-minute units! By doing so, you’ll accelerate the speed of learning, and the time it takes to reach your desired level will be shorter. Read more here: Study tip: Learn in 10-minute units!
After 20 minutes, we forget 40% of what we have learned. One hour later we already forget half. And after one day, more than 70%. These numbers come from Ebbinghaus. So repeat everything (every lesson, every de-coding, etc.) at least three times to store the majority in long-term memory. Read more here: Repeat it to keep it: Third Time’s a Charm!
3. Breaks & Sleep
It sounds strange, but it actually helps. A night of healthy sleep is extremely important for better memory. Studies, such as from the University of California, show that those who lie down for 20 minutes at noon increase their thinking ability compared to non-sleepers. During a nap, information from the day is processed, and important information is transferred from short-term memory to long-term memory.
Additional tip: MOVIE language courses for entertainment while learning.
With the Brain-Friendly language courses, you learn in a brain-suitable way: vocabulary is developed in context, and grammar is learned intuitively. It is as easy as child’s play, just like our mother language back then. The completely new learning concept offers carefree learning with a great fun and entertainment factor. Take a look!