Artificial Intelligence, Robots

How is the learning process of an Artificial Intelligence?

The majority of living beings that exist in the world go through a growing and learning process in which someone else teaches them how to do several things. If you look at humans, when we are children, our parents teach us specific behaviors and how to accomplish our obligations. If you see animals, their parents teach them how to hunt, survive and to defend themselves. We can find the same connection with Artificial Intelligence, there is always going to be someone teaching, or submitting information.

The evolution of the Artificial intelligence can be explained through faces. The first face consists in teaching the robots to do the best way possible a specific task. For example, if you want them to learn how to work in customer service, they are going to obtain the information from a group of people who are actually offering that service and they are going to absorb all their knowledge in order to store it in the neural network of the robot, meaning its brain.
The second face will be to expand their capacity of doing more than one task, which is the objective of the Artificial Intelligence in a long term and its whole purpose of existing.

What AI developers seek from the Artificial Intelligence is to make a simulation on how the learning process works in a human brain. You train them to do a specific task, and once they have done it in the correct way, they receive a reward. The same happens with human beings. When someone teaches us how to make a certain activity and we do it how it is supposed to be done, they give us a reward in order for us to know that it is the correct way to do it. The same way happens with the Artificial Intelligence. You give them rewards in order to keep them motivated to reach the goal: doing their tasks the best way possible, and in case they do it wrong, the programmers teach them the correct way to do it, so they can try it again. This is a try and error cycle in which it keeps repeating itself until it accomplishes the answer expected. This process takes a lot of time, but practice, perseverance and meticulousness are key to make this job perfect and worth it.

It is very important to highlight that, what is crucial at the moment the programmers teach these robots, is going to be the data they are going to store into their neural network, because they are not going to have the notion of determine what is good or what is bad. This happens with human beings too. We know how to differentiate between good or bad because our parents taught us how to with experiences, examples and talks. It’s the same with the Artificial Intelligence. They are going to learn what the creators want them to learn and their purpose is to be the best on those tasks.

There are people who are afraid of our future with this advanced technology. They ask themselves if we are going to be safe in a world with robots who are being educated to overcome human intelligence. But we need to understand that their behavior doesn’t depend on them, but from their creators and the information they want them to store. There is no point in between, the Artificial Intelligence is going to learn the information the programmers filter them with. Besides, there is going to be one point in which the computer is going to need the help of a human being because their knowledge is not infinite… Or at least for now.