Artificial Intelligence: Definition, History, and Future

7.1 The Classical Definition of AI

Since their emergence, computers have brought revolutionary changes to various aspects of human activity, generating significant interest in the development of intelligent machines. This interest, rooted in history, gave rise to the concept of artificial intelligence (AI), an idea that remains the subject of active discussion and development.

The earliest references to the creation of machines capable of imitating human intelligence can be found long before the advent of modern computers. The idea of artificial intelligence was inspired by the desire to expand human capabilities through technology, which gave rise to many philosophical and scientific studies. But what do we mean by the term “artificial intelligence,” especially when even the definition of the word “intelligence” itself appears difficult?

Defining “intelligence,” just like defining “artificial intelligence,” is a complex task subject to numerous interpretations. Over the decades, scholars have proposed different definitions of AI, which can be grouped into several main categories:

AI as a field of computer science: Defined as the development of computer systems capable of performing tasks that require human intelligence, including learning, reasoning, and self-correction.

AI as the improvement of machine functions: Treats AI as a process of improving machines so that they can perform tasks traditionally associated with human intelligence, such as adaptation and self-correction.

AI as an extension of human intelligence: Views AI as a tool for enhancing human intelligence, similar to the way mechanical tools enhanced physical strength.

AI as the study of effective programming methods: Focuses on developing techniques that allow computers to solve problems more efficiently.

Over time, definitions of AI have evolved along with technological progress. Early definitions focused on imitating human cognitive processes, whereas newer approaches emphasize the idea that AI should not merely copy human behavior, but should also strive for rationality and efficiency in problem solving, possibly even surpassing human abilities in some respects.

Modern AI research covers a wide range of approaches, including machine learning, deep learning, neural networks, and much more, each offering unique methods and techniques for creating intelligent systems. Among the key tasks of AI is the development of algorithms that can adapt and learn from data, providing new opportunities for automation and the solution of complex problems in various fields, from medicine to space research.

There are many ways to classify AI systems; however, most of them can be reduced to four main categories:

Systems that think like humans: This implies the development of AI systems that imitate human thinking and cognitive processes, such as perception and reasoning. The key goal here is to understand the human mind and create models that can reproduce aspects of human thought.

Systems that act like humans: Here the emphasis is placed on reproducing the external manifestations of human intelligence. Examples include systems capable of recognizing speech and gestures, understanding natural language, and even displaying emotions.

Systems that think rationally: This approach focuses on logical “thinking” or the use of rules to infer new information from known data. Systems developed within this direction seek to use formalized principles of logic to imitate rational thinking similar to human thinking.

Systems that act rationally: This involves creating systems that act in an optimal way to achieve a specific goal. They do not merely imitate human behavior, but also perform tasks efficiently and rationally under different conditions. This approach includes reinforcement learning and agent-based learning, where systems learn to optimize their behavior through a series of trials and errors.

7.1.1 Modern Interpretations and Areas of AI

Artificial intelligence (AI) is not merely a term, but also an actively developing scientific discipline that undergoes many changes and adaptations over time. Its modern definitions represent a combination of methods, technologies, and principles aimed at creating machines capable of imitating, or even surpassing, various aspects of human intelligence.

The following are the main areas and methods actively used in modern AI:

Deep Learning: A technology based on multilayer neural networks. This method stands at the forefront of many breakthroughs in image processing and speech recognition. Example: facial-recognition systems in social networks or photography assistants in smartphones.

Reinforcement Learning: A methodology in which algorithms “learn” in practice by choosing actions that produce maximum benefit in the long term. Example: algorithms that learn to play chess or video games, achieving high performance.

Cognitive Modeling: An approach in which models are created that imitate human cognitive processes, including reasoning, learning, and memory. Example: systems that model decision-making under uncertainty.

Robotics and Autonomous Systems: Thanks to AI, new-generation robots can independently analyze their environment, make decisions, and act. Example: robot vacuum cleaners that clean a home independently, or self-driving cars.

Natural Language Processing (NLP): This area of AI allows machines to understand, interpret, and generate human language. Example: voice assistants such as Siri or Google Assistant, as well as automatic translation systems such as Google Translate.

Computer Vision: A scientific field that allows machines to “see” and interpret visual information, similarly to human perception. Example: QR-code scanning systems or automatic diagnostic systems for medical images.

Neural-Network Modeling: Studies the brain in order to create algorithms that imitate human cognitive functions. Example: creating neural networks that can classify and analyze complex data patterns.

It is also worth noting that, as AI develops, emphasis on its ethical and social aspects is increasing. Issues of safety, transparency, and accountability are becoming more and more relevant. This field continues to expand, incorporating new methods and approaches, which makes it one of the most dynamic areas of modern technology.

7.1.2 Key Characteristics and Properties of AI

Artificial intelligence has a number of unique characteristics that distinguish it from other technologies and make it a subject of intensive research. The main characteristics and properties of AI are as follows:

Learning capability: AI algorithms can analyze data, adapt to new information, and improve their performance over time without direct human intervention. This property underlies most AI applications, allowing systems to improve and evolve independently.

Autonomy: Allows systems to make decisions independently on the basis of previous learning and available data, which is especially valuable in conditions where human participation is limited or impossible. This ability to operate autonomously makes AI indispensable in tasks such as controlling autonomous vehicles or monitoring remote systems.

Processing large volumes of data: The ability of a system to quickly analyze and interpret enormous datasets, identifying patterns and correlations in them that are inaccessible to human analysis. This ability to analyze big data makes it possible to apply AI to trend prediction, process optimization, and support for data-driven decision-making.

Multitasking and adaptability: Allows the same algorithms to adapt and perform different tasks depending on context and changing conditions.

Reasoning and planning: Advanced systems can process information logically, draw conclusions, and plan actions several steps ahead.

Interaction with natural language and perception of the surrounding world through sensor data allow AI systems to interact with people more naturally and understand the world around them. These capabilities make AI more intuitive and accessible to a broad audience, opening new paths for humanizing technological interactions.

Imitation of human emotions: Some advanced AI systems can recognize and even imitate human emotions, making their interaction with people more natural.

Self-awareness: Although full-fledged self-awareness for machines is still a subject of science fiction, some researchers are seeking ways to develop systems that could possess some level of “consciousness” or self-understanding.

7.1.3 The Difference Between AI, Automation, and Traditional Programming

In the modern world, the term “artificial intelligence” has become almost commonplace; however, not everyone understands exactly how AI differs from ordinary automation or traditional programming. At first glance, it seems that all these concepts mean the same thing: machines performing certain actions instead of a human. But on closer examination, the key differences become visible.

Traditional programming is based on creating clearly defined algorithms. A programmer gives the machine specific instructions on how to respond to different situations or input data. Every step and every action of the machine is predetermined by a human, and the computer simply follows these instructions. If situations arise that were not anticipated by the programmer, the program will not be able to handle the task or may even fail.

Automation, in turn, is the process of simplifying or accelerating tasks that were previously performed manually, using mechanisms or programs. This may include everything from automatic temperature control in a home to mass production at an enterprise. But again, such systems operate within the framework of clearly specified rules.

Artificial intelligence, by contrast, is based on the principle of learning. Instead of defining every possible situation in advance, AI is trained on the data with which it interacts and makes decisions on the basis of its experience. This allows systems to adapt to new situations, improve their effectiveness, and even “think” creatively in some situations. Such systems are capable of identifying patterns in data that may be invisible to humans and, on that basis, making predictions or taking actions.

Artificial intelligence does not simply replace automation or traditional programming; it provides tools for solving problems that previously seemed impossible. However, this also implies greater responsibility, because decisions made by AI may have far-reaching consequences in the real world.

7.2 The History of the Development of Artificial Intelligence

Artificial intelligence, despite being perceived as a relatively new area of technology, has roots that reach deep into antiquity, when philosophers first began asking questions about the essence of reason and the possibility of imitating it. These ancient reflections on the nature of intelligence and logic played a key role in forming the foundations for the development of the concept of artificial intelligence, although at that time these ideas seemed more like philosophical reflections than real technological possibilities.

Ancient philosophers such as Aristotle asked fundamental questions about what makes the human mind unique and considered whether it might be possible to create something similar to human intelligence, or even to surpass it. These questions became the starting point for reflections on logic, which Aristotle and other thinkers of that time viewed as a means of rational thinking and argumentation. They investigated the structure of logical reasoning, trying to understand how inferences are formed and follow one another, thereby laying the foundation for understanding rational thought in a broader sense.

As mathematics and science developed in subsequent centuries, these ideas began to take more concrete forms. Mathematical and logical systems developed by scientists became the basis for the first attempts to formalize reason and logical processes. These attempts at formalization underlie the modern algorithms and mechanisms used in artificial intelligence. Scientists and engineers began to develop machines and programs capable of performing logical operations and imitating aspects of human thinking, relying on mathematical principles and logical structures laid down by scientists of the past.

7.2.1 The Origins of Artificial Intelligence

The foundation of modern research in artificial intelligence was laid by the work of outstanding scientists, among whom Alan Turing occupies a special place. His concept of the “Turing machine” not only provided a theoretical basis for the development of computing systems, but also inspired the creation of machines capable of imitating aspects of human thinking. Turing also proposed the famous “Turing Test” as a methodology for assessing a machine’s ability to demonstrate intelligent behavior comparable to that of a human.

At the dawn of computer technology, the potential of computing systems to imitate the human mind was already evident to scientists and engineers. This period was characterized by the active development of research aimed at creating programs capable not only of solving mathematical problems and playing chess, but also of demonstrating elementary forms of learning. These early achievements opened the door to the further development of artificial intelligence, leading to the creation of systems that today can perform much more complex and diverse tasks.

Nevertheless, questions about the nature of consciousness and the possibility of its full imitation by machines remain a subject of philosophical debate. On the one hand, some researchers and philosophers are skeptical of the idea that machines will ever be able to fully reproduce all aspects of human intelligence. On the other hand, many believe in the limitless possibilities of artificial intelligence and its potential to reach, and even surpass, human intelligence. These debates remain relevant, forming the philosophical and ethical context within which modern AI develops, and they will undoubtedly influence its future direction.

7.2.2 The Era of the First Computers and the Birth of AI

The birth of artificial intelligence is closely linked to the development of the first computers. Although the ideas of automation and machine computation had existed for a long time, it was the emergence of the first electronic computing machines that became the starting point for the active study of the potential of machine intelligence.

With the appearance of the first computers, scientists received a tool that could perform complex calculations at a speed never seen before. This opened the door to the implementation of many theoretical ideas that had previously seemed impossible. Interest in the capabilities of these machines was so great that many saw them not merely as calculators, but also as potential “thinking” machines capable of competing on equal terms with human intelligence.

At this time, the first research groups began to form that dealt with questions of artificial intelligence. Scientists from different fields, whether mathematics, psychology, or neurobiology, joined forces to understand how it might be possible to create a machine capable of “thinking.” They sought ways not only to increase the computational power of computers, but also to give them the ability to learn, reason, and perhaps even create.

One of the milestones in the development of artificial intelligence was the creation of programs that could play chess, solve mathematical problems, and even understand natural language, although only at a very primitive level. These achievements were striking because they demonstrated the potential of machines in fields traditionally considered the prerogative of human intelligence.

Nevertheless, researchers encountered many obstacles along the way. Limited computational resources, a lack of suitable data for training, and an incomplete understanding of the mechanisms of the human brain made the task of creating artificial intelligence extremely difficult. But even these obstacles could not stop scientists who were convinced that machines could, and eventually would, think.

As computer technology and algorithms developed, artificial intelligence became increasingly complex and advanced. And although many of the early ideals and ambitions were not achieved during that period, they laid the foundation for future achievements and continue to inspire researchers around the world.

7.2.3 The Golden Age of Artificial Intelligence: 1950–1970

The period from 1950 to 1970 is often referred to as the “Golden Age” of artificial intelligence, marked by extraordinary optimism, breakthrough scientific discoveries, and active funding of research in this field. During these years, the fundamental principles and concepts of AI were laid down, and they continue to define the direction of development of this discipline.

The early 1950s were marked by intense interest in the possibility of creating machines capable of reproducing human intelligence. This interest was fueled by advances in cybernetics and the emergence of the first computing machines, as well as by philosophical and scientific reflections on the nature of intelligence and the uniqueness of the human brain.

Alan Turing, one of the key figures of that time, made a significant contribution to the development of this field by proposing the “Turing Test,” a method for evaluating a machine’s ability to demonstrate intelligent behavior comparable to human behavior. This test became a catalyst for numerous studies aimed at developing machine intelligence.

During these two decades, the first programs and prototypes were created that demonstrated elements of intelligent activity, such as the ability to play chess, solve complex logical problems, recognize patterns, and interact in natural language. These achievements received significant recognition and support, stimulating further development in the field of AI.

However, along with progress, it became evident that the path toward creating full-fledged artificial intelligence was more intricate and difficult than had been assumed at the outset. Many early approaches proved unable to scale to solve more complex tasks, raising doubts about their long-term viability.

Despite these obstacles, the “Golden Age” of artificial intelligence left an indelible mark on the history of the discipline. It laid the theoretical and practical foundations on which modern research is built and inspired numerous innovations that shape today’s AI world.

7.2.4 The AI Winter: 1970–1990

The period from 1970 to 1990 in the history of artificial intelligence is often called the “AI Winter,” symbolizing the lull that followed the rapid growth and enthusiasm of the preceding decades. This stage was characterized by a slowdown in research, reduced funding, and growing skepticism about the attainability of the ambitious AI goals that had seemed to be on the horizon during the “Golden Age.”

The United Kingdom became the scene of a key event of this period in 1973 with the publication of the “Lighthill Report,” which critically assessed the current state of AI research, especially emphasizing the absence of significant achievements in the areas of machine translation and automatic speech recognition. This report led to a reduction in government funding in the United Kingdom and became a precursor to similar measures in other countries.

Almost simultaneously, in the United States, the Defense Advanced Research Projects Agency (DARPA) significantly reduced funding for AI projects, justifying its decision with similar arguments about a lack of progress and concrete results, which further intensified the effect of the “AI Winter.”

Nevertheless, this period was not without significant events and achievements. In 1980, the expert system XCON was developed for Digital Equipment Corporation, demonstrating the successful commercial application of AI by helping specialists configure and assemble computing equipment. Despite limited capabilities and high development and maintenance costs, expert systems became an important step in the commercialization of AI.

In addition, research in neural networks continued during this period, although not as intensively as before. An important achievement was the development of the backpropagation algorithm in 1986, which helped accelerate the training of multilayer neural networks and renewed interest in this direction.

Overall, however, these two decades left behind many questions and doubts about the further development of artificial intelligence. Many scientists and researchers felt disappointed because they did not see the expected breakthrough, and investors and policymakers became more cautious in their expectations. Despite this, it was precisely these difficult times that prepared the ground for the next rise of artificial intelligence, which began in the 1990s.

7.2.5 The Revival of Interest in AI: 1990–2010

Beginning in the 1990s, the world witnessed a revival of interest in artificial intelligence. This period was marked by a number of breakthroughs in science and technology, as well as the return of active funding for research in this field. Neural networks once again became the focus of attention, undergoing a revolution thanks to improved learning methods and the growth of computing power.

In the mid-1990s, one notable event was the emergence of machine-learning algorithms based on decision trees, such as random forest. These methods offered new approaches to data analysis and prediction, and since then they have been actively used in various fields. For example, in 1996 IBM introduced its chess program Deep Blue, which the following year, in 1997, defeated world chess champion Garry Kasparov. This was not only a triumph of technology, but also proof of AI’s potential in solving complex problems.

While progress in neural networks was noticeable, their true potential became evident only with the development of deep learning in the 2000s. Improved training methods, combined with the exponential growth of computing resources and the availability of enormous volumes of data, led deep neural networks to begin demonstrating striking results in tasks such as image or speech recognition. For example, in 2009 Google began developing self-driving cars, using AI technologies to process sensor data and make decisions in real time.

By 2010, accumulated scientific and technological achievements in AI had led to a revival of belief in its potential. Startups, research laboratories, and large corporations actively invested in the development of AI-based products and solutions. Thus, in 2011 Apple introduced its virtual assistant Siri, which became a symbol of a new era in human-machine interaction.

In conclusion, the period from 1990 to 2010 was a time of rethinking and intensive development of artificial intelligence. These years serve as a bridge between the previous decades of searching and the modern AI boom that we observe today.

7.2.6 The Deep Learning Boom and the Modern Period: From 2010 to the Present

Since 2010, the world of artificial intelligence has been experiencing a true renaissance, thanks to a surge of interest in deep learning. Deep learning, based on multilayer neural networks, has caused revolutions in a number of industries, from medicine to the automotive industry.

One of the key moments confirming the potential of deep learning was the ImageNet image-recognition competition in 2012. A neural network called AlexNet, developed by researchers from Toronto, reduced the recognition error almost by half compared with previous methods. This success became a catalyst for applying deep learning in many applications.

The following years were marked by a number of important events demonstrating the power of deep learning. In 2015 Google introduced its DeepMind’s AlphaGo system, which was able to defeat the world champion in the game of Go, a game considered one of the most difficult for machines. This moment became a kind of “Sputnik moment” for the field of artificial intelligence, because it showed that machines can surpass human abilities even in the most complex intellectual tasks.

At the same time, deep learning became a driver of innovation in natural language processing. Automatic translation services such as Google Translate began actively using deep learning technologies to improve translation quality. In 2018 OpenAI introduced the GPT-2 model, which is capable of generating texts practically indistinguishable from human-written ones.

In addition, since 2010 artificial intelligence has come to play a central role in the development of autonomous vehicles. Testing self-driving cars on roads has become commonplace, and many automakers, such as Tesla, have begun introducing autonomous functions into their cars.

Deep learning has also brought revolutionary changes to medicine. AI-based systems have begun helping physicians diagnose diseases, analyze medical images, and predict potential risks for patients.

In conclusion, the last decade has witnessed remarkable progress in artificial intelligence. Deep learning, which became the true trigger of this progress, has allowed AI to find applications in almost every corner of our daily lives, changing familiar things and providing new opportunities for scientific and technological discoveries.

7.2.7 Key Figures and Their Contributions to the Development of AI

In the history of the development of artificial intelligence, several key figures can be identified whose contribution to this field was truly revolutionary. One such person is Alan Turing. As early as 1950, he proposed the concept of the “Turing Test,” an experiment that would make it possible to determine whether a machine has the capacity for human-like thinking. He suggested that if a machine can conduct a dialogue in such a way that a person cannot distinguish it from another person, then such a machine can be considered “thinking.”

Another notable scientist is John McCarthy. In 1956, he organized the first conference on artificial intelligence, which became the starting point for the formation of AI as a separate discipline. McCarthy also coined the term “artificial intelligence” and developed the LISP programming language, which became a standard for creating AI systems for many years.

Marvin Minsky is another legendary figure in the world of AI. Having founded, together with John McCarthy in 1959, the Artificial Intelligence Laboratory at the Massachusetts Institute of Technology, he conducted numerous studies aimed at creating machines capable of human-like thinking.

Raymond Kurzweil is also worth mentioning; he became known for his predictions about the future of AI. In 1990, in his book The Age of Spiritual Machines, he predicted that by 2029 machines would be able to pass the Turing Test, and that by 2045 the so-called “technological singularity” would occur, when the intellectual capabilities of machines surpass those of humans.

There are also figures such as Yann LeCun, who made a significant contribution to the development of convolutional neural networks, which became the basis for progress in computer vision. In 2013, he joined Facebook to head a new AI research division.

In addition to the scientific leaders mentioned above, many other researchers have also made invaluable contributions to the development of artificial intelligence.

Geoffrey Hinton, often called the “godfather of deep learning,” played a key role in popularizing deep-learning algorithms, in particular backpropagation. Beginning in the 1980s, his work in this area became the foundation for subsequent innovations in deep learning. In 2013, together with Yann LeCun and Yoshua Bengio, he became one of the key figures in the “deep learning boom,” when corporations and scientific institutions began investing heavily in this field.

Yoshua Bengio, a professor at the University of Montreal, also made a significant contribution to deep-learning research, and his works are often cited as foundational in this field. Thanks to his research, many new architectures and methods for training neural networks appeared.

Sergey Brin and Larry Page, the founders of Google, not only created one of the most powerful search engines, but also actively invested in the development of artificial intelligence. Their decision to acquire DeepMind in 2014 showed the whole world their belief in the potential of deep learning. In subsequent years, Google became one of the leaders in AI, developing products such as Google Assistant and AutoML.

Rosalind Picard, a professor at the Massachusetts Institute of Technology, is also worth mentioning; she was at the origins of “affective computing,” an area of AI that studies the ability of machines to recognize and respond to human emotions.

Each of these individuals, by making a unique contribution, helped shape and develop artificial intelligence into what we know today. But the list certainly does not end there. Many researchers, engineers, and enthusiasts around the world continue to work on creating new technologies and improving existing ones, making the world a more advanced and interactive place.