In the first part of my series of articles, we clarified the basic concepts of artificial intelligence and its contemporary significance. However, it is\u00a0Development of AI<\/strong>\u00a0It is not the result of a technological explosion in a vacuum, but of a centuries-old intellectual journey. A\u00a0History of Artificial Intelligence<\/strong>\u00a0It is about one of humanity's deepest ambitions: Not only do we want to understand our own cognitive processes, but we also aim for their artificial reproduction and technological realization.<\/p>\n\n\n\n This endeavor is deeply rooted in the co-sciences. Philosophy has been researching the origins of knowledge for thousands of years, mathematics has defined the boundaries of formal systems, in particular through Kurt G\u00f6del's 1931 non-completion theorem, and economics has contributed to the model of rational agents with the fundamentals of game theory and decision theory. AI research builds on these interdisciplinary foundations to bridge the gap between biological intelligence and silicon-based logic.<\/p>\n\n\n\n The scientific dawn of artificial intelligence dates back to 1943, when Warren McCulloch and Walter Pitts published their groundbreaking study. The revolutionary realization of the authors was that neurobiology, formal logic, and the Turing machine theory were common denominators. In their model, biological neurons were treated as simple logic gates (\u2018on\/off\u2019 state units).<\/p>\n\n\n\n It has been shown that such a network of two-state neurons can, in principle, perform any predictable arithmetic and Boolean function. This demonstrates that biologically inspired structures can cover the full spectrum of digital logic. Donald Hebb provided a theoretical framework for the learning process in 1949, pointing out that the key to learning is to modify the connections between neurons, called weights. This principle led to the construction of the first network simulator, the 40 neurons SNARC, in 1950.<\/p>\n\n\n\n \u201cDuring learning, the strength of connections between neurons changes: links that are often activated together are strengthened.\u201d \u2013 Donald Hebb\u2019s rule, which is the direct theoretical ancestor of weight correction algorithms (e.g. backpropagation) in modern neural networks.<\/p>\n\n\n\n In his visionary writings (1947, 1950), Alan Turing, the father of modern computer science, transcended mere computational capacity. His paradigm-shifting idea was a metaphor for the \u2018child-machine\u2019: Instead of trying to rigidly program a static, adult-level intelligence, we need to create a flexible basic system that can learn and evolve from its environment.<\/p>\n\n\n\n Turing also defined the criteria that are essential for a machine to be considered intelligent. To successfully complete the famous Turing test, the machine must have four basic capabilities: natural language processing (NLP), knowledge representation (storage), automated inference and machine learning.<\/p>\n\n\n\n AI as an independent discipline was born in the summer of 1956 at a two-month workshop at Dartmouth College. John McCarthy was not only the host of the event, but also the creator of the term \u2018artificial intelligence\u2019. The conference brought together the different scientific approaches of the Founding Fathers (\"The Four Great Ones\"):<\/p>\n\n\n\n The optimism of the participants was limitless: It was believed that a decade or two would be enough to artificially map the basic functions of human intelligence. Although their predictions were exaggerated, this event triggered the first institutionalized wave of AI research.<\/p>\n\n\n\n Since the 1970s, research has turned to expert systems. These programmes tried to replicate the knowledge of human experts in a narrow field (e.g. medical diagnostics or technical configuration) by building on the \u2018if-then\u2019 type of logic rules.<\/p>\n\n\n\n This was the first era of great business success for AI. The most famous example is Developed for DEC R1 (XCON)<\/strong> a system that automates the compilation of computerised configurations. The system Saving $40 million a year<\/strong> brought to the company in 1986. During the dominance of the knowledge-based approach, the AI industry grew into a billion-dollar market, demonstrating for the first time the economic viability of the technology.<\/p>\n\n\n\n Throughout history, the development of AI has stalled several times. The \u2018MI winters\u2019 of the 1970s and 1990s were the result of a gap between excessive expectations and real technological constraints. The reasons for the depletion of resources were several:<\/p>\n\n\n\n In 1997, IBM Deep Blue<\/strong> His computer brought AI back to public consciousness when he defeated World Chess Champion Garri Kasparov. Although the success of Deep Blue was largely due to the brute force method and pre-programmed chess skills, the symbolic significance of the victory was enormous.<\/p>\n\n\n\n This period marked a shift from rigid rule-based systems to Machine Learning (Machine Learning)<\/strong> towards. The focus has shifted from predefined rules to pattern recognition: the systems no longer simply implemented the instructions, but began to extract correlations from the data themselves.<\/p>\n\n\n\n In the 2000s the Big Data<\/strong> and a GPUs<\/strong> (graphic processors) parallel processing ability brought the Deep Learning (Deep Learning)<\/strong> age. Multilayer neural networks have also become able to recognize complex, nonlinear relationships.<\/p>\n\n\n\n The technological leap is AlphaGo's 2016 victory<\/strong> It's against Lee Sedol. In Go, there are more combinations of steps than there are atoms in the observable universe, so the raw computational force would have failed here. The success of AlphaGo was the triumph of intuitive pattern recognition and learning from parties played against itself.<\/p>\n\n\n\n Deep Blue (1997) vs AlphaGo (2016) Technological Differences:<\/strong><\/p>\n\n\n\n Modern large language models (LLMs), such as ChatGPT, represent the pinnacle of probability pattern recognition. It is important to understand: these systems do not \u2018think\u2019 in the human sense of the word, but reflect the statistical context of the input data.<\/p>\n\n\n\n Let's take the the Apple Color<\/strong> an example of pattern recognition:\u00a0If AI has seen 10 examples in the course of teaching, of which 7 times the apple is red, 2 times green and 1 times yellow<\/strong> It was then that the next word was replaced by the highest probability (70%) is assigned to the word \u2018red\u2019. This statistical prediction provides the basis for the \u2018intelligence\u2019 of modern AI.<\/p>\n\n\n\n The current technological spectrum:<\/strong><\/p>\n\n\n\n The path of AI is continuous evolution from logical gateways to self-learning networks:<\/p>\n\n\n\n In the next part of my series of articles, we will look in detail at the different types of AI and find out where the boundary lies between Weak AI, Strong AI and Superintelligence in the distant future.<\/p>","protected":false},"excerpt":{"rendered":" Welcome to this series of articles! As an expert and technology educator, my goal is to help dispel the mysticism and possible anxieties surrounding artificial intelligence (AI). <\/p>","protected":false},"author":2,"featured_media":1557,"menu_order":0,"comment_status":"closed","ping_status":"closed","template":"","format":"standard","meta":{"slim_seo":{"facebook_image":"https:\/\/teszarypeter.hu\/wp-content\/uploads\/2025\/10\/AI-Category-Cover.jpg","twitter_image":"https:\/\/teszarypeter.hu\/wp-content\/uploads\/2025\/10\/AI-Category-Cover.jpg","title":"2. R\u00e9sz: A logika \u00fatj\u00e1n \u2013 Az MI t\u00f6rt\u00e9nete Turingt\u00f3l napjainkig - Tesz\u00e1ry P\u00e9ter","description":"\u00dcdv\u00f6z\u00f6llek ebben a cikksorozatban! Szak\u00e9rt\u0151k\u00e9nt \u00e9s technol\u00f3giai eduk\u00e1tork\u00e9nt az a c\u00e9lom, hogy seg\u00edtsek eloszlatni a mesters\u00e9ges intelligenci\u00e1t (MI) \u00f6vez\u0151 miszti"},"footnotes":""},"categories":[93],"tags":[94,96,95],"class_list":["post-2116","publikacio","type-publikacio","status-publish","format-standard","has-post-thumbnail","hentry","category-ai","tag-ai","tag-mesterseges-intelligencia","tag-mi"],"desktop_mode_lock":null,"desktop_mode_contributors":[],"desktop_mode_attached_media":[1557],"meta_box":[],"_links":{"self":[{"href":"https:\/\/teszarypeter.hu\/en\/wp-json\/wp\/v2\/publikacio\/2116","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/teszarypeter.hu\/en\/wp-json\/wp\/v2\/publikacio"}],"about":[{"href":"https:\/\/teszarypeter.hu\/en\/wp-json\/wp\/v2\/types\/publikacio"}],"author":[{"embeddable":true,"href":"https:\/\/teszarypeter.hu\/en\/wp-json\/wp\/v2\/users\/2"}],"replies":[{"embeddable":true,"href":"https:\/\/teszarypeter.hu\/en\/wp-json\/wp\/v2\/comments?post=2116"}],"version-history":[{"count":3,"href":"https:\/\/teszarypeter.hu\/en\/wp-json\/wp\/v2\/publikacio\/2116\/revisions"}],"predecessor-version":[{"id":2256,"href":"https:\/\/teszarypeter.hu\/en\/wp-json\/wp\/v2\/publikacio\/2116\/revisions\/2256"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/teszarypeter.hu\/en\/wp-json\/wp\/v2\/media\/1557"}],"wp:attachment":[{"href":"https:\/\/teszarypeter.hu\/en\/wp-json\/wp\/v2\/media?parent=2116"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/teszarypeter.hu\/en\/wp-json\/wp\/v2\/categories?post=2116"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/teszarypeter.hu\/en\/wp-json\/wp\/v2\/tags?post=2116"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}2. The Dawn: McCulloch, Pitts and Biological Inspiration (1943)<\/h3>\n\n\n\n
3. Alan Turing and the Machine That Thinks<\/h3>\n\n\n\n
Concept<\/th> The idea of Turing<\/th> Modern equivalent<\/th><\/tr> Turing test<\/strong><\/td> It's a communication test where the interviewer can't decide if he's talking to a human or a machine.<\/td> Evaluation of LLMs (e.g. GPT-4) and competitions for the Loebner Prize.<\/td><\/tr> Children's machine<\/strong><\/td> Creation of a basic learning system instead of programming the finished knowledge.<\/td> Deep Learning and Reinforcement Learning.<\/td><\/tr> Genetic algorithms<\/strong><\/td> Development of algorithms following the principles of evolution (selection, mutation).<\/td> Evolutionary strategies for optimizing network architectures.<\/td><\/tr><\/tbody><\/table><\/figure>\n\n\n\n 4. 1956: The Official Birth - The Dartmouth Conference<\/h3>\n\n\n\n
\n
5. The Golden Age and Expert Systems (1970s-1980s)<\/h3>\n\n\n\n
6. AI Winters (AI Winters): Periods of Disappointment<\/h3>\n\n\n\n
\n
7. The Renaissance: Deep Blue and the Return of Machine Learning<\/h3>\n\n\n\n
8. The Deep Learning Revolution and AlphaGo (2000-present)<\/h3>\n\n\n\n
\n
9. Today's AI: The power of ChatGPT and pattern recognition<\/h3>\n\n\n\n
\n
10. Summary<\/h3>\n\n\n\n
\n