The 11 types of artificial intelligence (and their characteristics) - medical - 2023
- What is artificial intelligence?
- How is artificial intelligence classified?
- 1. Reactive machines
- 2. Theory of Mind
- 3. Self-aware machines
- 4. Limited memory
- 5. Narrow artificial intelligence
- 6. General artificial intelligence
- 7. Artificial superintelligence
- 8. Systems that act like humans
- 9. Systems that think like humans
- 10. Systems that act rationally
- 11. Systems that think rationally
Intelligent robots have been the engine of both film and literary plots for many years.. In these works, science fiction was nourished by futuristic scenarios where machines had an intelligence comparable to or superior to that of human beings, with positive or negative consequences depending on the work in question.
Who was going to tell us that, in so few years, we would get this to stop being science fiction and become science to dry? It is incredible how technological progress has allowed the development of machines and computer systems that, although not intelligent in the strict sense of the word, have cognitive capacities capable of imitating those of the human being.
We live surrounded by algorithms capable of perceiving, reasoning, learning and solving problems. Nobody knows how far artificial intelligence can go, but what we do know is that, although it is still in very premature stages, it will lead (if it has not already led) to a revolution equal to or more important than the Internet itself .
And in today's article, then, we will clear up all the doubts you may have about artificial intelligence. What exactly is artificial intelligence? Can a machine be intelligent? It's dangerous? What types are there? What differentiates from each other? Get ready to dive into this exciting world of robots and intelligence.
- We recommend reading: "Will the day come when the human being can be immortal?"
What is artificial intelligence?
"Artificial intelligence" is a difficult concept to define, although roughly it can be understood as intelligence carried out by machines. Unlike natural intelligence, it does not involve consciousness or emotions. But let's go further.
Artificial intelligence refers to any computing device capable of perceiving its own environment and responding with specific actions in order to maximize the probability of meeting its objectives.. Whatever they are. From knowing which ads to show us when we surf the net, to clinical programs to detect cancer in patients, to strategy games such as chess against a computer or the promising future of autonomous cars.
In this sense, artificial intelligence is the computing consequence of using algorithms designed for the purpose of creating machines with cognitive capacities comparable to those of the human being. And, although it seems somewhat distant, it is day by day in our life: Google, social networks, emails, electronic commerce (Amazon uses artificial intelligence to recommend products based on your user profile), online entertainment services (such as Netflix or Spotify) , GPS applications, Medicine, mobile facial recognition, chats with robots ...
We live surrounded by increasingly intelligent machines and immersed in computer worlds that, for better or worse, determine our lives. Artificial intelligence is a branch of computational science that studies how computer programs are capable of providing machines with abilities typical of human beings., with an important weight in two properties: reasoning and behavior.
Therefore, artificial intelligence looks for programmed strategies based on algorithms (or the famous artificial neural networks) that allow machines to imitate the functioning of human neurons, thus endowing these systems with the ability to solve problems as a human being.
But what exactly are these algorithms? Basically an algorithm is a set of written instructions that the machine must follow to execute certain commands. Any computer system is based on algorithms. It responds to certain commands, but in a linear fashion. And this does not make you smart.
Therefore, artificial intelligence is possible thanks to what are known as intelligent algorithms, which allow the machine to respond to commands but not in a linear way. That is, the algorithms themselves give the system the ability to interpret situations and data, responding differently in each case. Logic and mathematics. This is what cool but efficient artificial intelligence is based on.
A machine has artificial intelligence only and exclusively when the algorithms programmed into it allow it to develop calculations not only to respond to commands, but also to learn from the situations to which you are exposed, since each calculation that it performs results in information that it stores for future situations.
- We recommend reading: "Science and cinema: is there real science in movies?"
How is artificial intelligence classified?
The term "artificial intelligence" was first coined in 1956 by John McCarthy, Claude Shannon and Marvin Minsky, although for more than fifteen years this discipline was not investigated, as the scientific community believed that it was something typical of science fiction.
Much has happened since then and, today, systems that make use of artificial intelligence are here to stay, both for better and for worse. But are all artificially intelligent systems the same? No. Not much less. There are different types of artificial intelligence that we will analyze below.
1. Reactive machines
Reactive machines are those systems endowed with the most primitive artificial intelligence. They are machines capable of mimicking the human mental ability to respond to stimuli, but do not have the ability to learn. In other words, they cannot use previous experience to develop more effective responses. An example would be the famous Deep blue from IBM, a reactive machine that, in 1997, was able to beat Garry Kasparov, the Grandmaster of chess, in a game of chess.
2. Theory of Mind
The Theory of Mind is a concept that is not yet implemented but is being worked on, as it is the next level of artificial intelligence. A system with this level would be able to understand those entities with which it interacts by being able to interpret such complex and subjective information as needs, beliefs, thoughts and even emotions. They will be machines capable of understanding humans. In other words, we will be facing an artificial emotional intelligence.
3. Self-aware machines
According to specialists, this would be the last level of artificial intelligence. When machines were aware of their own existence, artificial intelligence would have reached its peak, since the system would have emotions, needs and even wants. And here come the apocalyptic scenarios. What if this machine had the desire to annihilate us? Either way, we are decades (and even hundreds of years) away from having the technology to create self-aware systems. Artificial self-awareness is, for now, a hypothetical concept.
4. Limited memory
Limited memory machines are those that have the ability to respond to stimuli from reactive machines, but we add a very important component: learning. Systems with this artificial intelligence are able to learn from actions in order to memorize effective ways to respond to certain situations. The artificial intelligence that surrounds us is of this type. From Google algorithms to self-driving cars.
5. Narrow artificial intelligence
Better known by its English name, Artificial Narrow Intelligence (NAI), narrow artificial intelligence refers to all those computational systems that, despite imitating human capabilities, are designed to perform only a specific task. They cannot do anything other than what they are programmed to do. They have a very narrow range of activity, hence their name. For now, all the artificial intelligence implemented in the world is of this type.
6. General artificial intelligence
Better known also by its name in English, Artificial General Intelligence (AGI), general artificial intelligence will be (not yet developed) a form of artificial intelligence that will allow machines to form connections between different activities. You will be able to learn, understand, act and perceive the environment just like a human being. They will not be limited to a single activity, but will be able to imitate our abilities to learn everything. With training, a general artificial intelligence will be able to increase its range of action more and more.
7. Artificial superintelligence
Another hypothetical concept. With artificial superintelligence, better known as Artificial superintelligence (ASI), machines will not only imitate the way of thinking of human beings, but will become, by far, the most intelligent system on the face of the Earth. They will be better than us at absolutely everything. They will learn everything at unimaginable speed and the door to uniqueness will open.; a situation in which artificial intelligence will go through an intellectual explosion in which it itself will generate more and more powerful machines. And so on until who knows where. A really scary scenario that, luckily, is far from materializing.
8. Systems that act like humans
Systems that act like humans are all those systems that They do not imitate our way of thinking, but rather that of behaving. It is not intended that they develop complex cognitive abilities, but rather that they perform mechanical actions more effectively than we do. Robots are a clear example of this.
9. Systems that think like humans
Systems that think like humans are all those systems that do they imitate our way of thinking. This artificial intelligence is equipped with algorithms that allow it to learn, memorize, make decisions and solve problems. It is the most prevalent form of artificial intelligence.
10. Systems that act rationally
The systems that act rationally are all those systems that imitate humans when it comes to rational behavior. They perceive the environment and, from there, carry out actions as effective as possible to meet their objective.
11. Systems that think rationally
The systems that think rationally are all those systems that imitate humans, but not in mechanical or learning actions, but in terms of rational thinking. They are machines that have, in logic, the pillar of their operation. Through calculations, they reason and then act.