The rise of Artificial Intelligence (AI) in business is very topical. Its use is spreading and is changing, even, the models [...]Read More »
Unlike a computer program, in which a list of orders are processed through a computer program, AI goes beyond the mere execution of a set of orders to obtain a result and is capable of learning to perform more complex tasks by imitating mechanisms of human learning, in a way AI makes computers program themselves. We can find different types of Artificial Intelligence according to their capacity and functionality.
The proper functioning of AI is largely subordinated to the algorithms and processes used to develop it. In this sense, one can speak of weak and strong AI systems.
In the case of weak or narrow AI, we are talking about a system designed and trained to perform a single task. On the other hand, when we refer to strong AI, also called general AI (AGI - Artificial General Intelligence), we are talking about a system that has cognitive abilities similar to human ones, which allows it to find the solution to a given task on its own.
To answer the question What is artificial intelligence? We should, first of all, delimit the scope or types of artificial intelligence to which we are going to refer. Broadly speaking, AI is the ability of a machine or computer system to emulate intelligent behavior and thought, learning and creation processes comparable to those of a biological human being, i.e., capable of analyzing the environment and performing actions, with a certain degree of autonomy, to achieve specific objectives.
"A computer can be called 'intelligent' if it manages to fool a person into thinking it is a human" (Alan Turing).
The development of AI is based on processes that include learning, reasoning and self-correction. The more these processes are "internalized" in the machines, the greater their capacity for self-management, although there is still a long way to go before we can talk about fully autonomous AI systems and much further to go before they become self-aware.
Currently, AI is present in many fields and its development is growing exponentially. We can find specific applications of AI in a multitude of sectors, such as healthcare, human resources, commercial-marketing, security, risk or video games. Current AI systems are capable of handling large amounts of data, extracting relevant information and creating profiles or predictive models in many fields.
► You may be interested in: Main applications of AI in enterprises
One of the most important innovations brought about by the development of AI is the automation of repetitive processes in companies. AI is also proving vital to economize resources and time spent for data analysis, strategy definition, etc. Let's look at some examples.
"Automation will reduce the time to perform tasks that are impossible for humans to do today ... The amount of data that companies have is beyond the capability of the most sophisticated data scientist or professional. But that's not true for machine learning ... The ability to turn that into information to help you sell more and help you save more - AI will affect both."(Mark Hurd, Oracle CEO)
As a result of digital transformation, AI applied to the exploitation of available data is significantly improving the performance of many businesses. The number of businesses using AI in one of their processes has increased by 270 % in the last 4 years.
► You may be interested in: Artificial intelligence in business: the secrets of some successful examples
We can find more and more fields of application for AI. This is having a great impact on companies. In fact, the development of AI for innovation in large companies is fundamental.
From product recommendations while browsing the internet to the operation of smartphones themselves. In addition, AI is closely linked to social demands.
In the following, we will analyze what are the types of Artificial Intelligence that we currently find around us and those that are yet to land in the future.
This is one of the most common types of Artificial Intelligence. ANI is an AI with a reactive character, limited memory and oriented to specific objectives. Its function is to be prepared to act in a single role, ensuring that it fully performs its role.
This category includes everything from virtual assistants such as Alexa or Siri to autonomous vehicle systems, as well as email spam filters or advertising recommendations based on our searches. In short, Narrow AI is based on implementing a series of actions and instructions in the machine.
Of the different types of Artificial Intelligence, this type of AI, from a theoretical point of view, is capable of performing any type of task with the same effectiveness as a biological human being. In other words, it is a cognitive AI or, in other words, it has a personality, although it is still at a very early stage of development. Unlike the Narrow AI is about simulating in the machine all the human brain processes for autonomous decision making in different scenarios.
This implies that AGI has the ability to evaluate and detect different needs, processes and even emotions in order to act accordingly. For example, strategies can be modeled from the most common interactions, behaviors, doubts and needs of a platform's users.
Although it is unknown today whether this degree of complexity can be achieved, super AI should be able to perform any activity better than a human being, without necessarily replicating their behavior.
Therefore, of the different types of Artificial Intelligence, this one would have the ability to think, reason and apply its own judgments to complex issues in a conscious and autonomous way. And even more: planning based on experience, learning and communicating on its own.
Some researchers believe that it is possible to reach this level of technology since the brain is a mechanical system, it should be possible to simulate it using synthetic materials, although the complexities of human thought suggest both a physical and biological impossibility.
In addition, the advanced reasoning capabilities of humans require meticulous analysis of the potential environmental and ethical consequences this may have on future machines.
Science fiction movies are full of robots and machines that reach a technological maturity unthinkable today. This is the case of films such as 'Blade Runner', 'Terminator', 'Her' or '2001, A Space Odyssey'.
At this point we focus on the usefulness of Artificial Intelligence in real, not hypothetical, domains.
This is the most basic AI, in which a certain range of responses (reactions) is implemented for specific requests. Therefore, the machine has only one response role, i.e., it is automated to react to a given action on the present.
Nor can they use previous experiences on which to base current decision making because it does not have the ability to learn and manage an internal database to run what it absorbs and evolve.
One of the classic examples in the development of this type of machine was the "Deep Blue" supercomputer, created by IBM in the 1990s, which defeated the best chess players in the world, including Garry Kasparov.
Deep Blue was able to recognize figures on a board and process 200 million moves in one second. However, it had no concept of past actions, i.e., it ignored any data, relevant or not, before the present moment.
In this case, the machine, which is still completely reactive, is able to store past experiences or learn from recent data for a short and limited period of time. This allows it to generate timely actions from the information gathered and add it to its programming to create new patterns of behavior and response for the not too distant future.
Also, as in the case of reactive machines, no learning is generated based on experience. An example of this type of AI is autonomous cars, which analyze the speed and direction of other vehicles.
Machines using this type of AI are able, based on their history of interactions, to make timely decisions to respond to a request or perform an action. By extending the memory of these machines, promising results are being achieved in, for example, chatbots and facial recognition systems.
In the field of psychology, the so-called 'theory of mind' implies that human beings possess thoughts and emotions that affect their own behavior and define what we know as social interaction.
Therefore, if machines are to walk among us, they must have an understanding of how we think and how we feel. They will also have to come to know what we expect and how we want to be treated. They will have to adjust their behavior accordingly. This brings us to the next level of AI.
In the future, machines will be self-aware. That is the path that some scientists have set as a horizon, although concrete prototypes have not yet been developed.
There is currently no machine that is self-aware. At the computational level, this is the most ambitious goal facing AI. A self-aware machine would be able to store past data and, based on it, create its own judgment and act accordingly.
The problem is that the AI would have to include in its programming the ability to understand that there are individuals with emotions and thoughts of their own, which would be a level of complexity unattainable for current computer systems.
"The only limit to AI is the human imagination" (Chris Duffey).
This type of AI could understand the world around it if fully developed. Therefore, machines would be able to learn based on our behaviors and deduce and know what our tastes, needs, desires or even how we expect to be treated.
Obviously, much of the future of AI is framed around the development of machines capable of developing self-awareness without the limitations of biology. This would imply that machines will develop self-awareness and be able to recognize themselves as independent entities. However, it is not possible to say when, or even if, the technology will be able to advance to that level.
We often wonder what examples of AI we can find in our environment, and the fact is that artificial intelligence is a concept that in English has [...]Read More »