History of Artificial Intelligence (AI)

Pioneers of AI Evolution (Created with AI | Dall-E 3 Plugin in ChatGPT)

AI as an Idea | 322 BC - 1818

Who would have thought that? The history of Artificial Intelligence did not start in big tech companies. Instead, it was born a long time ago in the minds of philosophers and writers. Even ancient philosophers were already thinking about machines that could work without human assistance. Aristotle was one of the first to think about the idea of automated mechanisms. His ideas formed the intellectual foundation for today’s AI concepts. In his work “Politics” he wrote:

"[...] if every tool could perform its own work when ordered, or by seeing what to do in advance, […] master-craftsmen would have no need of assistants and [...] of slaves."

Aristoteles

In the centuries that followed, curiosity and creativity for machines that make people’s lives easier developed into a global phenomenon. In ancient China, the engineer Yan developed a life-size, human-like robot to impress King Mu of Zhou. In Jewish literature, the artificial creature Golem was brought to life. This was usually a clay-bound creature that had no free will but was strong and powerful. Around 1300, the Arab intellectual and engineer Al-Jazari created the elephant clock, in which mechanical figures woke up every half hour to make sounds.

Even the great visionary of the Renaissance, Leonardo da Vinci, made sketches of mechanical figures whose complex movements were far ahead of their time, although they were never built.

And Paracelsus described an artificially created human being, the homunculus, in his work “De natura rerum”.

Last but not least, a significant contribution to the development of AI has been made by science fiction literature. In “Gulliver’s Travels” Jonathan Swift introduced “Engine”, the first computer-like machine, in 1726. , Also Mary Shelley dealt in “Frankenstein” with the subject of artificial life.

Thus, the concepts of Artificial Intelligence that are currently in the headlines of the daily press did not suddenly arise. They were built on a rich heritage of stories, speculation and creative thoughts, preparing us to explore the deeper layers of Artificial Intelligence.

The Birth of AI | 1950s - 1970s

While the early concepts and visions of Artificial Intelligence set a solid foundation and shaped the dreams of machine intelligence, in the 1950s to 1970s these dreams moved closer to reality.

“Specify the way in which you believe that a man is superior to a computer and I shall build
a computer which refutes your belief.”

Alan Turing

Alan Turing was at the heart of this revolutionary phase. Turing was a pioneer. In his paper “Computing machinery and intelligence”, he revolutionized the way we think about the intelligence of machines. Turing raised the question: “Can a machine think?” In order to approach this question, he developed the “Turing test” named after him to test whether a machine can think or communicate in such a way that it is impossible to distinguish it from a human being. Turing thus created the starting point for a new field of scientific AI research.

Just six years later, in the summer of 1956, an important founding moment of artificial intelligence took place at Dartmouth College. John McCarthy was one of a small group of visionary thinkers. McCarthy introduced the term “Artificial Intelligence” for the first time. It was the moment when AI became a research area in its own right, exploring the potential of machines that could not only think, but learn independently – which was a revolutionary idea at that time.

At the same time, during the same meeting at Dartmouth College, Allen Newell and Herbert A. Simon developed a program called “Logic Theorist”. This program made history as the world’s first real AI program. It was able to solve several dozen mathematical problems, demonstrating the impressive capabilities of computers. The success of the “Logic Theorists” did not only demonstrate the capabilities of AI research, but also the enormous potential that lies in the machine imitation of human brain.

Just ten years later, Joseph Weizenbaum set another milestone in research into communication between humans and machines by developing the first chatbot in history, called “ELIZA”, at the Massachusetts Institute of Technology (MIT). ELIZA simulated a psychotherapist who could respond logically and relevantly to the input from a chatbot. Using amazingly simple techniques, ELIZA succeeded in giving human users the impression of an empathic conversation partner. Weizenbaum was surprised and at the same time concerned about the users’ trusting nature. With this impressive simulation of human-like abilities, Weizenbaum triggered a fundamental debate about the societal, ethical and social dimensions of human-machine interaction. Questions about the potential and limits of AI arose, as well as about the responsibility of imitating human thought and interaction.

Between Hope and Disappointment: AI Winters and Expert Systems | Mid 1970s - 1990s

The emergence of programs like ELIZA gave great optimism about what AI could achieve. However, limitations of this technology back then, particularly the limited computing power and memory capacity, soon made clear that the road to creating an intelligent machine was far more complex than anticipated. In the 1970s, the AI world experienced its first major downturn.

In his Lighthill Report, Michael James Lighthill criticized the fact that progress in AI research fell short of the ambitious promises. He pointed out that – despite large investments – the results remained disappointing, particularly in the areas of general problem solving and language comprehension. He also pointed out that increasing complexity was making it difficult to fully grasp the potential of Artificial Intelligence. As a result, he made a recommendation for cutting financial support for AI projects. Faith in intelligent machines suffered. The research community experienced the first great AI winter: A period of disappointment followed, resulting in significant cuts in research funding.

Of course, AI researchers were not completely put off. At the same time, expert systems experienced a breakthrough. In 1972, ‘MYCIN’ was developed at Stanford University to advise doctors on the selection of suitable antibiotics. Six years later, ‘XCON’ was created at Carnegie Mellon University to automate complex computer configurations at Digital Equipment Corporation – a task that previously required the expertise of experienced technicians. XCON saved the company up to 40 million dollars a year, which resulted in a worldwide demand for expert systems from companies in the 1980s, thus becoming the main focus of AI research. These advanced computer programs were designed to simulate human expertise in specialized fields such as medicine, chemistry or engineering and demonstrated for the first time how Artificial Intelligence could overcome everyday challenges. Once again, Artificial Intelligence achieved significant successes and marked the beginning of a new era.

However, the development and maintenance of expert systems was challenging. They required regular updates in order to keep pace with the continuous growth in specialist knowledge. In addition, expert systems were often unable to meet the high expectations placed on their performance and flexibility. These problems, coupled with the high costs of the necessary hardware and software development, once again led to growing skepticism and disappointment. Interest and investment in AI research declined once again. Finally, in 1987, the second AI winter began with the failure of the high expectations placed on expert systems.

The Awakening of AI: Breakthroughs and Innovations that Are Changing the World | 1990s - Today

The history of artificial intelligence is characterized by breakthroughs and setbacks. But every time AI research has reached its limits, the AI community has used these findings to learn from past mistakes and move forward with renewed enthusiasm. Since the late 1990s, the amount of data from the Internet and the performance of computers have increased significantly. A new era of Artificial Intelligence began:

The triumph of the IBM computer Deep Blue over the world chess champion Garry Kasparov brought AI back into the public spotlight during this period. Once again, this moment showed that Artificial Intelligence has the potential to exceed human experts in complex tasks. Interest in AI research, which had stagnated since the second AI winter, came back to life.

In 2012, Alex Krizhevsky, Ilya Sutskever and Geoffrey Hinton sparked a groundbreaking revolution in the field of deep learning with the development of their deep neural network, which they named AlexNet. AlexNet’s significance was shown in its outstanding performance at the ImageNet Challenge on September 30, 2012, a competition that aimed to encourage advances in the field of image recognition and classification. AlexNet outperformed the competition by far, achieving a significantly lower error rate in image recognition than leading models so far. AlexNet impressively demonstrated the potential of deep learning, thereby catalyzing a significant increase in research activity and innovation in the field of artificial neural networks. As a result, deep learning models have made significant advances in a wide range of application areas, including speech recognition, text processing and autonomous driving. Deep learning has established itself as a central component of modern Artificial Intelligence.

AlphaGo’s triumph over Lee Sedol in March 2016 was not only another milestone in the field of Artificial Intelligence, but also proof that AI can mimic human intuition in one of the world’s most complex board games. Go, a game considered significantly more complex than chess due to its enormous number of possible moves and strategic depth, posed a particular challenge. Traditional approaches had not been sufficient to cope with the complexity of Go. AlphaGo learned by studying millions of Go games and exceeded all expectations. The brilliant idea of using neural networks and self-learning processes goes back to the mathematician Irving John Good in 1965. However, it was the development of AlphaGo by the Google DeepMind team that made it possible to successfully implement this approach. With consequences: AlphaGo’s victory boosted investment in AI research worldwide. Immediately after the match, the South Korean government announced that it would invest one trillion won (around 863 million US dollars) in AI research.

The Future of Artificial Intelligence | 2024 - Future

The recent era of AI, characterized by the development and application of Large Language Models, such as Google Duplex and ChatGPT, has once again revolutionized the field. These models have been trained on large amounts of data and demonstrate human-like language processing capabilities.

In May 2018, Google Duplex caused a sensation and opened up a new era in human-machine communication. The Google Assistant’s advanced technology independently arranged an appointment with a hairdresser over the phone without the other person realizing they were talking to an AI.

The unveiling of ChatGPT by OpenAI on November 30, 2022 caused a worldwide sensation. Within just two months, ChatGPT became the fastest growing software application in history with over 100 million users. With its ability to respond to complex queries in a way that is almost indistinguishable from human interaction, ChatGPT immediately attracted attention and sparked a flood of wonder, admiration and debate. ChatGPT demonstrates previously unimaginable versatility and competence: it solves complex math problems, writes creative texts and simulates in-depth conversations. Researchers at Stanford University confirmed that ChatGPT-4 passes the Turing test and differs from average human behavior only in its increased willingness to cooperate. ChatGPT not only represents another milestone in AI research, but also marks another turning point in our understanding of Artificial Intelligence.

Today, Artificial Intelligence is everywhere and it is driving significant innovation in various industries. Advances in image and speech recognition have accelerated the integration of AI in numerous application areas, including autonomous vehicles, personal assistants and medical diagnostic tools.
AI systems can also revolutionize learning in the education sector, for example by creating personalized learning platforms. Such learning platforms can recognize the learning style and pace of individual students and customize the curriculum to make learning more personalized and efficient for students.

In customer service, AI-based chatbots and virtual assistants are revolutionizing the way companies interact with their customers. They answer inquiries, make bookings and provide important information, which significantly increases customer satisfaction and optimizes workflows.

In the creative industry, AI is used for music and film production as well as for the creation of computer games. AI systems not only accelerate the creative process, but also enable completely new forms of expression, such as algorithmically composed pieces of music and dynamically generated game scenarios. A recent study in “Scientific Reports” shows that AI systems have achieved higher average scores than humans in the Alternate Uses Task, a key creativity test, for the first time. Among these systems were OpenAI’s ChatGPT-3 and GPT-4, which set new standards in creative problem solving.

Current and future developments in AI technology are increasing productivity and efficiency. They also promise to profoundly change the way we work and interact with technology. This development brings with it both challenges and exciting opportunities for society.

The future will also be about AI applications that have the potential to fundamentally improve human experience and skills in an increasingly automated world and integrate them seamlessly into our everyday lives. The influence of AI is constantly growing. As AI becomes an integral part of our everyday lives, the challenge of using these technologies for the benefit of all is growing. In doing so, we will have to find answers to new questions regarding ethics, data protection and the future of work.

The history of AI should always serve as a reminder that all progress comes with both opportunity and responsibility. In this dynamic lies the true essence – a constant movement between exploration and reflection, aiming to push the boundaries of what is possible while always keeping the human at the center. The overarching goal should be to build a harmonious connection between humans and Artificial Intelligence.

By the way: As mentioned above, Google DeepMind’s AI defeated Go player Lee Sedol in 2016. Sedol ended his career in 2019, saying: “Even if I become the number one, there is an entity that cannot be defeated.” However, in February 2023 humanity struck back when Kellin Pelrine defeated the Go programs “Kata-Go” and “Leela Zero” The irony is that the strategy for victory was developed by an AI.

Did You Like Our Article?

Let us know what you think via our social media and stay up to date with us by following us and commenting on and sharing our content. We want to give you a comprehensive orientation about Artificial Intelligence and the future.

Category

Tag

From Aristotle to Algorithms | A Journey Through the History of AI