Artificial Intelligence (AI, for its acronym in English) is everywhere. On your screens, in your pockets and one day you can even walk to a house near you. This vast and diverse field is generalized into a single theme today.

https://i1.wp.com/criptotendencia.com/wp-content/uploads/2018/05/Inteligencia-Artificial-Robotica-1.jpg?w=1000&ssl=1
[Source](https://i1.wp.com/criptotendencia.com/wp-content/uploads/2018/05/Inteligencia-Artificial-Robotica-1.jpg?w=1000&ssl=1)

The robots that come out of laboratories, the algorithms that play games and win them, AI and their promises are becoming part of our daily lives. While all these instances have some relation to artificial intelligence, this is not a monolithic field, but one that has many separate and distinct disciplines.

Many times we use the term artificial intelligence as a general term that covers everything and that is not exactly the case. AI, machine learning, deep learning and robotics are fascinating and separate topics. All of them serve as an integral part of the greater future of our technology. Many of these categories tend to overlap and complement each other.

Artificial intelligence

At the root of IA technology is the ability of machines to perform tasks characteristic of human intelligence. These kinds of things include planning, pattern recognition, natural language compression, learning and problem solving.

There are two main types of AI: general and narrow. Our current technological capabilities fall under the latter. Narrow AI exhibits a splinter of some kind of intelligence, either reminiscent of an animal or a human. The experience of this machine is, as its name suggests, of limited scope. In general, this type of AI can only do one thing extremely well, such as recognizing images or searching databases at lightning speed.

The current AI technology is responsible for many incredible things. These algorithms help Amazon give you personalized recommendations and make sure your Google searches are relevant to what you're looking for. For the most part, any technologically literate person uses this type of technology every day.

One of the main differentiators between AI and conventional programming is the fact that non-AI programs are carried out through a set of defined instructions. AI, on the other hand, learns without being explicitly programmed.

This is when the confusion begins to take place. Many times, but not all the time, AI uses machine learning, which is a subset of the AI ​​field. If we dig a little deeper, we get deep learning, which is a way to implement machine learning from scratch.

Also, when we think about robotics, we tend to think that robots and artificial intelligence are interchangeable terms. AI algorithms are generally only one part of a larger technological matrix of hardware, electronics and non-IA code within a robot.

https://i2.wp.com/criptotendencia.com/wp-content/uploads/2018/05/Inteligencia-Artificial-Robotica-2.jpg?w=657&ssl=1
[Source](https://i2.wp.com/criptotendencia.com/wp-content/uploads/2018/05/Inteligencia-Artificial-Robotica-2.jpg?w=657&ssl=1)

It is a branch of technology that deals strictly with robots. A robot is a programmable machine that performs a set of tasks autonomously in some way. They are not computers nor are they strictly artificially intelligent.

Many experts can not agree on exactly what constitutes a robot. But for our purposes, we consider that it has a physical presence, is programmable and has a certain level of autonomy. Here are some different examples of some robots today:

 Roomba (vacuum cleaning robot)
 Surgery robots
 Atlas (humanoid robot)
Some of these robots for example, the robot of the assembly line or surgery bot are explicitly programmed to do a job. They do not learn. Therefore, we could not consider them artificially intelligent.

These are robots that are controlled by built-in AI programs. This is a recent development, since most of the industrial robots were only programmed to carry out repetitive tasks without thinking.

Self-learning robots with learning logic within them are considered IA. They need it to perform increasingly complex tasks.

 IA and Machine Learning

At its foundation, machine learning is a subset and a way to achieve true AI. It was a term coined by Arthur Samuel in 1959, where he stated: "the ability to learn without being explicitly programmed."

The idea is to obtain the algorithm to learn or be trained to do something without being specifically coded with a set of particular instructions. It is machine learning that paves the way for artificial intelligence.

Arthur Samuel wanted to create a computer program that would allow his computer to beat him in ladies. Instead of creating a detailed and lengthy program that could do so, he thought of a different idea.

The algorithm he created gave his computer the ability to learn, since he played thousands of games against himself. This has been the crux of the idea since then. In the early 1960s, this program was able to beat the champions in the game.

https://i0.wp.com/criptotendencia.com/wp-content/uploads/2018/05/Aprendizaje-autom%C3%A1tico-IA.jpg?w=1000&ssl=1
[Source](https://i0.wp.com/criptotendencia.com/wp-content/uploads/2018/05/Aprendizaje-autom%C3%A1tico-IA.jpg?w=1000&ssl=1)

Over the years, machine learning developed in several different methods. Those are:

Supervised
Semi-supervised
Unsupervised
Reinforcement
Of all of them, reinforcement learning is the most prominent. An example is the game of chess. Know a certain number of rules and base your progress on the final result of winning or losing.

Deep Learning

For an even deeper subset of machine learning comes deep learning. He has the task of problems much greater than the simple rudimentary classification. It works in the scope of the data quantities of the containers and reaches its conclusion without any prior knowledge.

If you were trying to differentiate between two different animals, you would distinguish them in a different way compared to normal machine learning. First, all the images of the animals will be scanned, pixel by pixel. Once it is completed, it will then analyze the different edges and shapes, classifying them in a differential order to determine the difference.

Deep learning tends to require a lot more hardware power. These machines that execute this are usually stored in large data centers. Programs that use deep learning essentially start from scratch.

Of all the AI ​​disciplines, deep learning is the most promising for a day creating a generalized artificial intelligence. Some current applications that deep learning has rejected have been the many chatbots we see today. Alexa, Siri and Cortana from Microsoft can thank their brains thanks to this ingenious technology.

There have been many seismic changes in the world of technology in the last century. From the computer age to the Internet and the world of mobile devices. These different categories of technology will pave the way for a new future where we will move from mobile devices to a first world of AI.