Dale Lane - Machine Learning for Kids

When my dad was a child, growing up in Illinois after the turn of the centurynot this century, mind you, but the previous oneradio was getting started, the zipper had just been invented, Charlie Chaplin was all the rage, Lindberg had not yet crossed the Atlantic, the Titanic sunk, the Model T was released, and most of the world was still being governed by various monarchies or vast, sprawling empires. When I was a child, growing up in the Great Plains of Texas, we had just landed humans on the moon, computers were huge, expensive beasts, the internet had not yet been invented, plastic was starting to become widespread, and talking to someone on the other side of the world was prohibitively expensive for all but a few. I spent many a lazy summer afternoon making gunpowder in my bedroom (sorry, Mom!), launching model rockets, and dreaming about going to the stars. I built my first computer from scratch when I was 12, and when I say from scratch, I mean literally from individual transistors, diodes, resistors, and transistors. Id read about this remarkable robot called Shakey by the good folks at the Stanford Research Institute, and I sat mesmerized in the movie theater watching HAL from 2001: A Space Odyssey. I knew then and there that I wanted to build computers that would take people to the stars.

And now, here I am, doing very much that.

Right now, the world is a very, very different place than what my father and I experienced growing up. In some ways its better, in others it is not, but one thing that has not changed is our ability to dream and to work to make those dreams become reality. Today, you can gain access to the worlds knowledge using a device you can hold in your hand. Its possible to talkand seesomeone on the other side of the world, or even in space, in real time. Cars are now electric, and some of them even drive themselves. Not only have we gone to the moon, but Mars is now populated entirely by robots, and we have sent spacecraft beyond our solar system. Thanks to computers, we can produce more for less, we make art that augments and even transcends reality, we have extended our bodies and our minds, and we study the cosmos in ways far deeper and broader than wed ever imagined.

The even more remarkable thing is that, owing to the fact that computers have gotten so inexpensive and their computational power so vast, almost anyone can learn how to use computers to build things that are limited only by our imagination. This is where computational thinking comes in, a way of approaching the world in terms of abstractions and algorithms that make it possible to turn our imagination into systems that run on our computers. To build a program that does your finances, or controls the heating in your house, or calculates the trajectory of a rocket flung into space, youll think mostly in terms of symbols. But, if you want to build a car that drives itself, a robot that interacts safely with humans, or a video game that gives challenging virtual opponents, youll need to use a different way of working with computers, and that is where artificial intelligence and machine learning come into play. From the time of Alan Turingone of the pioneers of computingcomputer scientists have tried to make computers that reason and learn and behave as humans, and recent advances in machine learning have made it possible for us today to build such things.