Artificial intelligence: Definition, relevance, types and most important applications
Today we hear a lot about the term artificial intelligence; Most of us link it to science fiction movies in Hollywood, and imagine robots that dominate the world, but actually artificial intelligence is not limited to robots; It goes into many of the applications we use every day without feeling, and according to a recent survey of more than 1,400 consumers around the world, 63% of people don’t really realize that they’re using AI technologies.
Definition of artificial intelligence “A.I.”:
Artificial intelligence stands for “A.I.” for: (Artifical Intelligence)؛ It is known by more than one definition, including:
Artificial intelligence is a branch of computer science involved in creating smart machines and devices.
It is also defined as simulating the human brain in performing some of its complex functions, such as: (Learning, planning, speech discrimination, problem solving and mental and logical thinking).
It’s the invention of intelligent machines and software.
the ability of digital machines and computers to perform certain tasks that emulate and resemble those of intelligent beings, and artificial intelligence aims to access systems that are intelligent and behave as humans do in terms of learning and understanding.
Artificial intelligence is introduced in many applications, such as: smartphones, computers, cars, robots, drones, simulations from one area of artificial intelligence, such as video games that are developed to be more realistic, and applications that help teach the language.
History of artificial intelligence:
The term artificial intelligence emerged in the fiftieth century, specifically 1950; When the scientist Alan Turing introduced the Turing test, which evaluates the intelligence of a computer, it classifies it as “smart” if it can emulate the human mind.
The first software using artificial intelligence was created one year after the Turing test, by Christopher Strachey, who was head of programming research at Oxford University; He was able to play checkers through the computer and develop them.
Anthony Ottenger of the University of Cambridge then designed a simulation experiment of human shopping in more than one store through a computer, which was designed to measure computer learning ability, and was the first successful experiment of what is known as “machine learning.”
In 1956 at Dartmouth College, the concept of artificial intelligence was officially announced by John McCarthy, who organized a two-month workshop, bringing together researchers interested in artificial neural networks, which did not lead to any innovations, but was useful as bringing together the founders of artificial intelligence.
After this workshop, research into artificial intelligence began heavily and research centres were established; These centers focused on devising systems that find solutions to problems efficiently, such as logic theory, which is the first software for artificial intelligence, and systems that are self-contained as GPS.
in 1979; The first computer marching vehicle, known as the Stanford, was built.
in 1997; The first computer managed to beat a human competitor in chess.
Advances in artificial intelligence science began at the beginning of the 21st century; interactive robots became available in stores; Instead, there’s a robot that reacts to different emotions through facial expressions, and robots that do difficult tasks like Nomad; It is responsible for searching and exploring remote areas in the Antarctic and for locating meteorites in the area.
Types of artificial intelligence:
Artificial intelligence is classified according to its capabilities, for the following types:
- Limited or limited artificial intelligence:
This species is the most common species of the day, meaning: Artificial Intelligence (AI), which performs specific and obvious tasks, such as self-driving cars, speech or image recognition software, and chess on smart devices.
- General artificial intelligence:
This species works in like-minded ways; It focuses on making the machine capable of thinking and planning on its own, in a manner similar to human thinking, but to date there are no practical examples of this type, what exists are only research studies that require much effort to transform them into reality, and the method of “artificial neural network” is one of the methods of studying general artificial intelligence; Because it’s about producing a neural network system for the machine, similar to the networks of the human body.
- Super Artificial Intelligence:
It is a species that exceeds the level of human beings, so that it can perform tasks better than a specialized human being, and super-artificial intelligence has many of the characteristics it must have; the ability to learn, plan, communicate automatically and make judgments, but the concept of super-artificial intelligence is still a hypothetical concept that does not exist in our present era.
artificial intelligence classifications by function:
By function, artificial intelligence is classified into the following types:
- Interactive machines:
Artificial intelligence is the simplest type of artificial intelligence; Because it interacts with current experiences, but lacks the ability to learn from past experiences, for example; Google’s AipaGo system.
- Limited memory:
This type can store previous experience data for a limited period of time, and the best examples are the self-driving system.
- Theory of mind:
This kind of intelligence is concerned with machine understanding of human emotions, interaction and communication with people, and there are currently no applications to this kind of intelligence.
- Self-perception:
The future expectations of artificial intelligence seek to reach this kind of intelligence, so that machines have self-awareness and special feelings, making them smarter than the human being, and it’s still on the ground.
- Learn the machine:
It is also a subfield of artificial intelligence to learn the machine, where the computer is able to learn on its own from previous experiences; He’s able to predict and make the right decision. “Arthur Samuel” was first introduced in 1959.
Artificial intelligence applications:
Artificial intelligence has many applications, most notably:
- Interaction with visual system:
It means artificial intelligence applications that can explain and analyze their entered images, such as facial recognition programs, and image analysis for location recognition.
- Interaction with handwriting:
Handwritten recognition applications; Whether it’s the process of writing on paper or on the screen of the device itself.
- Smart robots:
Intelligent robots do the work of humans, and are characterized by their ability to sense ambient factors such as light, heat, sound, and motion, through special sensors, and the most important feature of these robots is that they can learn from their past experiences and take advantage of mistakes.
- Interaction with Sound:
Some applications of artificial intelligence are used to listen to speech and understand its meaning, even if the sound is in an atmosphere of noise, vernacular or street language, examples of which include:
- Siri application:
It is an audio service that allows the user to handle the phone only through voice commands, running on Apple devices, where the voice converts to words and searches online to meet users’ requests.
- Amazon:
Through computer learning, artificial intelligence identifies and proposes goods of interest to consumers, and thanks to this program, more than a third of the site’s sales come from these recommendations. Amazon has been using artificial intelligence technology for more than 20 years.
- Netflix:
Netflix offers viewer suggestions based on its knowledge of what it likes and likes, thanks to artificial intelligence; Netflix analyzes the viewer’s favorite, and therefore suggests works and movies.
- Electronic games:
Electronic games are becoming very popular, and every day a new game is invented or developed for an old game, artificial intelligence systems are used; Where these games require strategic thinking, like poker and chess.
- Social media:
Social networking uses artificial intelligence applications; Like: Facebook to detect hacking of user photos.
Artificial intelligence applications:
Artificial intelligence has many applications, most notably:
- Interaction with visual system:
It means artificial intelligence applications that can explain and analyze their entered images, such as facial recognition programs, and image analysis for location recognition.
- Interaction with handwriting:
Handwritten recognition applications; Whether it’s the process of writing on paper or on the screen of the device itself.
- Smart robots:
Intelligent robots do the work of humans, and are characterized by their ability to sense ambient factors such as light, heat, sound, and motion, through special sensors, and the most important feature of these robots is that they can learn from their past experiences and take advantage of mistakes.
- Interaction with Sound:
Some applications of artificial intelligence are used to listen to speech and understand its meaning, even if the sound is in an atmosphere of noise, vernacular or street language, examples of which include:
- Siri application:
It is an audio service that allows the user to handle the phone only through voice commands, running on Apple devices, where the voice converts to words and searches online to meet users’ requests.
Through computer learning, artificial intelligence identifies and proposes goods of interest to consumers, and thanks to this program, more than a third of the site’s sales come from these recommendations. Amazon has been using artificial intelligence technology for more than 20 years.
- Netflix:
Netflix offers viewer suggestions based on its knowledge of what it likes and likes, thanks to artificial intelligence; Netflix analyzes the viewer’s favorite, and therefore suggests works and movies.
- Electronic games:
Electronic games are becoming very popular, and every day a new game is invented or developed for an old game, artificial intelligence systems are used; Where these games require strategic thinking, like poker and chess.
- Social media:
Social networking uses artificial intelligence applications; Like: Facebook to detect hacking of user photos.
Artificial Intelligence Pros:
Artificial intelligence has many advantages, including:
- Lack of emotions:
Artificial intelligence systems are characterized by a complete lack of emotion, unlike the human being, who is governed by his emotions and temperament, which affects his performance and decision-making; These systems operate in a rational manner, making their decision objectively and in a short time.
- Work continuously:
The machine can operate continuously without fatigue or boredom, and its ability to produce is constant regardless of working conditions, unlike the human being that is most affected.
- Facilitating daily life:
Artificial intelligence has provided us with many important applications, which have facilitated our lives in many respects, and the smartphone is the biggest proof.
- Recurrent actions:
Artificial intelligence systems can be used to do iterative work; That is, it requires the same mechanism at a time.
- Medical care:
Artificial intelligence systems provide medical care to humans; This is through surgery simulators, applications that help detect neurological disorders, and radiation surgery applications that help to eradicate tumors without harming surrounding healthy tissues.
- Huge volume of data:
Artificial intelligence systems can handle, process and store an enormous amount of data.
- Accuracy:
Artificial intelligence systems provide high accuracy, and reduce the margin of error in the execution of tasks.
- Hard jobs:
It uses artificial intelligence systems to do hard work that humans can’t do, like excavation, exploration of hard-to-reach places like the ocean floor.
- Advice and guidance:
Some artificial intelligence applications offer advice and advice to human users in certain areas, such as the medical field.
Artificial intelligence negatives:
The cost of designing, implementing and even maintaining AI systems is very high.
Artificial intelligence systems don’t understand human values and ethics. They only do what they’re designed to do without looking at what’s right and wrong.
Artificial intelligence systems cannot change or develop their work system on their own, if they receive the same data every time.
The inability of artificial intelligence systems to innovate and innovate, and to respond to changes in the working environment, such as the ability of humans to do so.
Reliance on artificial intelligence systems, rather than human, has left many workers out.
The importance of artificial intelligence:
- In the medical field:
One of the most important examples of the importance of artificial intelligence is in this area; Forecasting of Central Care Unit transfers; Artificial intelligence (AI) systems are used to convert patients to the central care unit by guiding doctors to the starting point of treatment, as the patient may sometimes be transferred to the intensive care unit in an ill-thought-out manner; This results in poor results, with artificial intelligence systems using patients’ medical records, laboratory results, and biomarkers to remedy the condition of patients before they deteriorate, and having to transfer them to the intensive care unit.
- In business:
Artificial intelligence enhances the capabilities and capabilities of companies, increasing business efficiency and speed of execution, and the number of people interacting with these businesses through the development of their tools and software, and also, today, many modern companies are relying on artificial intelligence systems to offer their services instead of traditional employees.
The Future of Artificial Intelligence:
Scientists today seek to develop artificial intelligence; To make more use of it in the future, to make our lives easier, they started today with smartphones, cars, and future access to smart-system homes.
Future perceptions of artificial intelligence systems include:
Entertainment; It’s possible that a human being can watch a film that selects his actors.
Future artificial intelligence systems are better able to protect individuals’ personal data from theft and hacking.
Future IQ systems can be able to take care of children or the elderly, do housework, and even dangerous work such as firefighting and mine clearance.
Self-driving cars can be fully arrived at, leaving the car to the artificial intelligence systems available in it, and here we note that self-driving cars already exist in our time, but they will be substantially available in the future.
The difference between machine and deep learning
The previous example shows that the main difference between deep learning and machine learning is that machine learning models become progressively better, that they always need human intervention to give them broad lines of how they learn from data, and that deep learning teaches the algorithm itself, without relying on human intervention.
For example, if the machine learning algorithm is taught to light a particular lamp when hearing the word “dark,” the algorithm will respond only when hearing the word. If the model receives data such as “I’m unable to see anything if the light is too dim,” here machine learning techniques will not respond, but deep learning algorithms can conclude that the meaning is the same, and then respond and light the lamp.
Machine Learning
Scientists have observed that to reach human intelligence is not the origin of mimicking that intelligence, but learning the mechanism of learning in humans and then striving for intelligence, as humans gradually learn to read and write, and to read from simple books to hard books, this is exactly the idea of machine learning, feeding algorithmically a lot of data and allowing them to know things.
For example, an algorithm was provided with a lot of data on financial transactions, and I tell it about deception and leave it to indicate fraud and even predict fraud in the future.
The term was first coined by American computer scientist Arthur Samuel, who successfully designed the “Checkers” algorithm, and whose algorithm beat the 1962 world champion, known to Samuel: Give computers the ability to learn without being explicitly programmed.
Algorithm training involves feeding it with vast amounts of data and allowing itself to configure and improve itself, but the popular definition of machine learning is: A technique for analyzing data and learning from that data, then applying what they learned to make an informed decision.
Machine learning algorithms require training in large amounts of data, and the more data they provide for the algorithm, the better.
Examples of machine learning uses
Machine learning performs a variety of tasks, affecting almost every industry from the search for harmful IT security programs to weather forecasting to stockbrokers seeking cheap deals.
Nowadays, many major companies use machine learning to provide users with a better experience, for example, using Amazon Machine Learning to make better recommendations on product selection to fashion designers based on their preferences, and Netflix uses machine learning to make better suggestions to users from a TV or film series or shows that they want to see them.
Companies such as Microsoft use predictive machine learning models to provide better financial forecasts, making predictions on financial entities by learning from historical trends and generating stock movement forecasts.
Deep learning
Although machine learning algorithms treat many problems, complex and complex problems have emerged that these algorithms cannot solve, and like some actions that humans do so easily as facial recognition, voice or handwriting, they are very difficult for machines.
So as long as machine learning is about imitating how humans learn, why don’t machines mimic the human brain in the way of thinking, and from there it was the idea of creating neural networks, and the neural networks that were designed to solve certain problems began, and showed a lot of promise and could solve some complex problems that other algorithms couldn’t address.
But no matter how many neural networks have evolved, it is currently difficult to access the efficiency of the human brain, and it contains 86 billion neurons.
In order to develop these neural networks, scientists formed several layers of these networks and linked them together in a way that resembles the structure of brain cells, and each of those layers performs a certain function, and this deep sequence of those networks was behind the nickname “deep” on those algorithms.
Perhaps the simplest definition of deep learning is that it is a set of algorithms that attempt to learn at multiple levels, a very small field of artificial intelligence based on artificial neural networks.
This type of algorithm has been devised to solve important problems such as image recognition, voice recognition, natural language processing systems, and many others.
How does deep learning work?
As we mentioned earlier, deep learning algorithms are inspired by the information processing patterns found in the human brain, and just as we use our brains to identify patterns and classify different types of information, deep learning algorithms can be taught to accomplish the same tasks on machines.
The brain usually tries to decode the information it receives, achieving this by marking and identifying elements in different categories, and whenever we receive new information, the brain tries to compare it to a previously known element and understand it, which is the same concept used by deep learning algorithms.
The Reason for Deep Learning
And deep learning has advantages that are more likely than machine learning, chief among which is that it is supported by a huge amount of data because it is so common, in addition to extraction advantage, in the sense that human intervention does not need to set the main rules of learning, for example, if you want to use the machine learning model to determine whether or not a particular image displays a car, we need humans first to identify the unique advantages of a car. (Shape, size, windows, wheels, extract these features and give them the algorithm as input data.
In this way, the machine learning algorithm will classify the image, and this in machine learning, the programmer must directly interfere with the classification process.
In the case of a deep learning model, the feature extraction step is completely unnecessary, the model will recognize these unique characteristics of the car and make the right predictions without the help of any human being.
In fact, this applies to every other task you’re going to do with neural networks, so only the primary data is given to the neural network, and the rest is done through the model.
Examples of deep learning uses
The Google Brian project, from Google, may be the most prominent example of deep learning at work, researchers, without identifying any cat identification information, have trained and taught the system itself through 10 million cat images taken from YouTube videos, and the system has successfully identified cat images without using tagged data.
Without deep learning, we won’t have self – driving cars or automated chat programs or personal assistants like Alexa and Siri.
The Google Transit application will remain primitive, and Netflix will have no idea what movies or TV series we like or don’t like, if we don’t use deep learning algorithms.
We can even go so far as to say that the new industrial revolution is driven by artificial neural networks and deep learning.
What’s next?
The progress made by researchers working on DeepMind, Google Brain, OpenAI and various universities in artificial intelligence research is accelerating. Artificial intelligence day after day becomes able to solve the most difficult, even difficult, problems of human energy, which means that artificial intelligence changes faster than its history, and predictions of its future soon become old.
There are different schools of thought about how people talk or how they see the future of artificial intelligence. There are some who believe that progress in artificial intelligence will continue apace and tend to think a lot about super artificial intelligence, and whether or not this is good for humanity.
Another group does not believe that artificial intelligence is making much progress compared to human intelligence, and they expect another winter for it, where funding will end because of generally disappointing results, as in the past, while other researchers who rely on experiments and research, and struggle with chaotic data, are betting that the future is bound to be for applications of AI in various practical and cognitive fields.
