Artificial Intelligence was something we were able to see only in sci-fi movies. At that point, all the people who were into science knew that there were enthusiasts actively trying to develop a sort of AI. Nevertheless, it seemed like a thing of the distant future that our grand-grand-grand children would be able to experience. Everyone was wrong.
AI is present nowadays in all spheres of life. You desktop computers, smartphones, tablets, TVs, Bookmaker bonus codes, and more are probably using some kind of AI.
So what does this have to do with anything related to running? Read on!
What is DeepMind?
DeepMind is a subsidiary of Google and one of their biggest projects so far. Essentially, they are teaching AI to learn things by itself and apply them right away. In other words, the company researches and builds various AI systems that are safe, secure and applicable in the real world. Their AI learns all the time and constantly enriches the portfolio of problems it can solve.
The Viral Video
Recently, DeepMind gained the attention of the entire world by publishing a very strange and somewhat uncanny video. In it, a stick figure that has only legs jumps over various obstacles, crouches, leans forward and backward, and more.
At first sight, the whole video is kind of hilarious, especially if you like goofy cartoons with stickmen. However, once you learn that the figure in the video is an AI, things get rather peculiar.
However, everyone experiences quite a shock when they learn the most important thing about the silly video — the AI in it learned how to walk, run, and overcome obstacles by itself. How is that even possible?
How Does DeepMind Work?
How has it learnt to run by itself? For us, humans, this is as confusing as the big bang theory. Everyone who saw the video was left mesmerized by the fact that the little stick figure with funny legs was a powerful artificial intelligence.
In order to understand how DeepMind learned to walk, run and do parkour, you need to know the basics of AI programming. First of all, AI is called artificial intelligence because it resembles human intelligence up to a certain extent. In other words, the way it learns things is somewhat similar to the way we learn. We gather data via our five senses, process them, and apply what we have learned. It’s a basic input-output thing.
AI works in a similar way. For starters, you need to feed it a huge amount of data. Next, AI processes it by recognizing certain patterns. One of the best cases of how AI recognizes patterns is visible in online registration security checks.
You probably stumbled upon security checks where you need to select all parts of an image where cars (or some other objects) are present. A certain area of AI called image processing had been fed millions of car images until it learned to recognize when there were cars in a picture.
The way DeepMind works is similar to a certain extent. It gets a lot of “food” in terms of obstacles, and it is not “shown” in any way how to walk, run or overcome them. It extracts various data, processes it, recognizes patterns, and, most importantly, applies that to overcome obstacles.
This technology can help us in every aspect of our lives, and it is extremely important to follow the progress and see where things are going to take off from here! DeepMind is at the forefront of AI research, and, so far, the company has achieved extraordinary results.