The Silicon Valley celebrates deep learning algorithms. Essentially, it's about memorizing through constant repetition. But learning should actually be about something else. Recently, I was playing soccer in the garden with my two-and-a-half-year-old neighbor. When I got tangled in the goal net, he rushed to help me and shouted, "Wait, I'll get the spreader!" This refers to a hydraulic cutting tool used by the fire department to free trapped individuals from crushed vehicles. At that moment, three things crossed my mind.
- First: What on earth is a spreader?
- Second: Will he really be able to free me with it?
- And third: What kind of parents does this little one have?
Well, they are with the professional fire department. But did they really show him pictures of spreaders, rescue shears, and fire engines for days until he finally learned them? Highly unlikely. Learning with Pictures Currently, the Silicon Valley is celebrating this form of learning. So-called deep learning networks are supposed to teach themselves how a car looks, as long as they are shown enough pictures of cars. The principle seems plausible: they are supposed to chew through the data over and over again until similarities and correlations emerge. If an image then shows four wheels and a steering wheel, it's 99 percent likely to be a car. Practice makes perfect. Don't we know this from our own lives? Entire school systems and employee training seem to be based on this learning principle. It's as if constant cramming of information is the best way of learning - just like the latest algorithms from Google and Facebook already do when recognizing faces or cat videos. How Machines Learn to Think
- Perceive
- Understand
- Act
The IT industry may be mistaken here! Because learning is all well and good - but understanding is better. Once we have learned something, we can unlearn it. But once understood, we cannot un-understand it. And we don't understand by constantly repeating the same thing - but by grasping the underlying meaning once. My young neighbor probably only heard the term "spreader" once or twice. But that was enough for him not to forget it (and now I know too). We do this constantly and understand at first glance what the words selfie, flexitarian, or Teuro could mean. To Understand, One Needs Peace Those who want to not only memorize the world but also understand it should not stoop to the level of computer algorithms. Because they calculate away constantly, without sense or reason. Our Learning Tips
- Plan
- Repeat
- Sleep
On the other hand, we take a break when we see something new - activating those brain regions that grasp the meaning of new information. Numerous studies show that these moments of peace are essential for understanding. Sometimes, it's enough to take a closer look once, without long cramming. Instead of constantly bombarding ourselves with current news and updates, it's okay to take a break from the flood of information now and then to understand it. What a beautiful resolution that is.
