AI has reached a point where it’s handling to write emails, create super realistic paintings, predict global stock market and even diagnose complex medical problems. It all seems to have arrived overnight. But the cutting edge technologies driving modern business have not arrived overnight. They are the culmination of a century long journey that has been traveled by a group of rather eccentric mathematicians, logicians and rebellious computer scientists.
The technologies currently driving the modern business world as well as our daily lives have not come about overnight.
The technologies behind current business practices and daily life activities have not appeared overnight. Rather, they are the culmination of a long history of work by a series of somewhat eccentric mathematicians, maverick computer scientists, and visionary logicians in a multi-generational relay. So, who created artificial intelligence?
So now that you know who created artificial intelligence, let’s go through the fascinating history of the field, learn about the true of AI, follow the path of how we got here, and see how your business can leverage this century-long evolution to stay ahead of the pack.
1. The Ancient Roots: Philosophical Foundations of Thinking Machines
There has long been a human fascination with the concept of an artificial being, often referred to as a robot. While electronic circuits are a relatively recent development, attempts to 'artificialize' human form or function stretch back centuries. The root of AI is, in fact, philosophical and mythological.
Greek Mythology: The idea of the automated artificial being is also evident in classical Greek mythology. Pygmalion made a statue called Galatea, which was brought to life by the goddess Aphrodite. Hephaestus the blacksmith made Talos a huge (around 10 feet tall) bronze man (half man, half automaton), which was left on an island to guard it against invaders. Hephaestus gave Talos instructions to obey, which enabled him to function as an automated artificial being.
Aristotle formalized human thought in the 4th century BCE by developing syllogistic logic, which is the first formal system of human thought. By organizing the rules for true statements about the world to be true the syllogism provides the basis for the rule-based programming that now powers most of the world’s computers.
To mechanize calculations: In the 17th century, thinkers such as Leibniz and Pascal (most famously) created mechanical calculators. Leibniz wanted to create a universal language of human reason—what he called the Calculus Ratiocinator—a means by which human argument could be translated into simple mathematical equations. He famously commented that if two philosophers got into an argument, then they should simply sit down with their calculators and say, 'Let us calculate!'
These thinkers, although they were able to lay the groundwork for the development of AI, did not have the technology available to put their ideas into practice. It was not until the 20th century that all of the missing components came together to form what we now know as artificial intelligence.
2. The Mid-Century Spark: Alan Turing and the Birth of Computing
So when are we going to talk about the inventor of AI? Well, it was a brilliant British mathematician—the late Alan Turing.
+-------------------------------------------------------------+
| THE TURING TEST |
| |
| +------------------+ +--------------------+ |
| | Human Evaluator | <---------> | Human Content | |
| +------------------+ +--------------------+ |
| ^ |
| | |
| +---------------------> | Machine (AI) Text | |
| +--------------------+ |
| |
| "If an evaluator cannot reliably tell the machine apart |
| from the human, the machine has passed the test." |
+-------------------------------------------------------------+
Alan Turing was a brilliant British mathematician who, during World War II, cracked the German Enigma code using a machine called Bombe, saving millions of lives. But that was only the beginning for Turing’s very inquisitive mind.
Turing wrote a philosophical paper in 1950 entitled Computing Machinery and Intelligence. In this work, Turing sets out the problem of artificial intelligence by posing a simple question: Can machines think?
The Turing Test
Since determining the point at which human thought ceases to occur was not feasible, Turing put forward an alternative strategy in the form of The Imitation Game, today more commonly known as the Turing Test.
The game proposed by Turing for determining whether or not a computer really could think was to be an Imitation Game, or what has become known as the Turing Test. It was very simple and consisted of a human questioner who was presented with two input channels, one connected to a human and the other to a computer. Both the human and the computer deliver their outputs via a third channel, which is connected to the questioner. The questioner then delivers his questions via this channel and gets his answers from it. If the questioner cannot, on the basis of the answers to his questions, tell which of the two input channels is connected to the human and which to the computer, then the computer has passed the test and can be said to be capable of thought.
Turing did not limit his musings to posing this very important question. He predicted that by the year 2000, there would be sufficient memory available on computers to play the Imitation Game so well that an average interrogator would only be able to correctly identify the human from which the questions were being originated less than 70% of the time. While Turing was a bit optimistic in regards to true conversational capability as opposed to overall objective intelligence, his work would become the definitive model of what to aim for in achieving the objective of achieving intelligent machines. Thus, Alan Turing became known universally as the Father of Modern Computer Science.
3. The Dartmouth Workshop (1956): Coining "Artificial Intelligence"
Alan Turing might have conceived the soul of AI, but it was a group of American scientists who could be said to have given it a birthplace.
Dartmouth Summer Research Project on Artificial Intelligence (1956)
├── Organized by: John McCarthy
├── Key Attendees: Marvin Minsky, Nathaniel Rochester, Claude Shannon
└── Core Achievement: Officially coined the term "Artificial Intelligence"
Dartmouth Summer Research Project on Artificial Intelligence (1956)
├── Organized by: John McCarthy
├── Marvin Minsky
├── Nathaniel Rochester
├── Claude Shannon
└── Core Achievement: Naming the field “Artificial Intelligence”
The proposal had to be submitted in order to get funding for the workshop. The core of the proposal consisted of one sentence: “To carry out further work on artificial intelligence.” This was the first time this term was used. McCarthy chose it in order to clearly distinguish his field of research from cybernetics, a discipline that was primarily concerned with analog feedback systems. Artificial Intelligence would be the field of research on digital computers, using purely logical and symbolic methods.
He was the father of AI because he named the field Artificial Intelligence and spent his life expanding on the subject. (In addition to being the founder of AI, he invented the programming language Lisp which dominated the field for decades).
The Founding Fathers of AI
McCarthy wasn’t the only one at Dartmouth. We saw the coming together of a remarkable group of pioneers which went on to dominate computer science for decades:.
- Marvin Minsky: A great cognitive scientist who established the MIT AI lab. Minsky saw the human brain as a complex biological machine, which he spent his career replicating in computers.
- Herbert Simon and Allen Newell: This great team came to Dartmouth with a program they had put together called the Logic Theorist. That software may have been the first AI program in the world; it was able to present complex math theorems through symbolic logic. Also, it did in fact prove the theorems that, at the time, great mathematicians had been working on for years.
- Claude Shannon: The "which" is very much at the root of information theory. What he did do was to put forth that digital data could be measured, compressed, and transmitted without error. At Dartmouth it was very infectious.
The attendees really thought that a group of top scientists spending a summer together could make great progress in developing machines that think, form ideas, and improve upon themselves. Also, they were sure human-level AI would be a generation away.
4. The Interactive Evolution of AI: Share of Innovation Over Time
In the digital age, the early positivity of the 1960s hit a brick wall of what we today term "computational limits." By the mid-1970s computers did not have the processing speed and memory to work out complex real world issues. What had been put forward as basic word games or simple math based programs did the bulk of the early work out but for the large scale issues related to language translation or computer vision we saw a need for a level of computing power that did not exist on our planet at the time. This in turn led to frustration from governments and academic institutions, which in turn saw in the First AI Winter (1974-1980) the almost total drying up of research funds.
5. The Rollercoaster of Progress: AI Winters and the Rise of Expert Systems
Early attempts to develop programs that enabled computers to process language, as illustrated by the interactive explorer, were hindered by the enormous computational demands that such tasks required.
The early AI programs were, however, unable to overcome the inherent computational limitations of the machines of the 1970s. To process even simple word puzzles or elementary mathematical games required a considerable amount of processing power and memory, whereas the computers available in the early 1970s were in no way capable of handling more complex tasks such as language translation or even vision, as these would have required processing power and memory many times greater than was available on Earth at the time.
A first AI winter (1974–1980) followed because of the lack of immediate results, while the investments of governments and academic institutions largely were spent. In consequence, the funding for research was almost cut off almost completely.
The Rise of Expert Systems (1980s)
AI experienced a new boom in the 1980s when corporations created a large market for expert systems.
For decades, the goals of developing AI were centered on the task of developing a broad, human-like general intelligence. However, Edward Feigenbaum and other pioneers found a different route to applying AI in very profitable ways. Hyper-specialized software programs, or expert systems, became the focus of a large, corporate-driven effort in the AI community. An expert system is a computer program that is loaded with a large amount of human expert knowledge. This knowledge is first encoded into a database of IF-THEN rules. The rules are then run through an inference engine, which applies the rules to the facts of a case in a logical order to come up with conclusions.
[Human Expert Knowledge] ---> [Encoded as IF-THEN Rules] ---> [AI Inference Engine] ---> [Business Solution]
For example, an expert system to be used for oil exploration might have rules from top geologists regarding properties of rock in different geological environments. A large oil exploration corporation could save hundreds of millions of dollars using such a system. These systems became very popular, and large amounts of commercial funds were invested in them, leading to the Second AI Winter (1987–1993).
By providing hundreds of millions of dollars in return on investment for corporations, expert systems elicited a massive commercial investment. These systems, however, were brittle, very expensive to update by hand, and tended to fail when faced with a single case outside of their rules. This led to the Second AI Winter (1987–1993).
6. The Modern Era: Data, Deep Learning, and Neural Networks
True world revolution is not governed by manual rules of programmers but by a learning computer. The true revolution of the modern world was sparked by this change of focus by the computer scientists. This is the story of the pioneers of the two fields of machine learning and deep learning.
Machine Learning and the pioneers of Deep Learning.
- The Rebirth of Neural Networks
- When a handful of researchers in the late 1980s and 1990s continued to develop artificial neural networks, the main tool for artificial intelligence, computer science experts from the mainstream scientific community threw them away as ineffective. Three of these pioneers have since earned their title of Godfathers of Deep Learning: Geoffrey Hinton, Yann LeCun, and Yoshua Bengio.
- While the mainstream scientific community was labeling neural networks as a dead-end approach that wouldn't be effective for years to come, three individuals, often referred to as the "Godfathers of Deep Learning," proved the skeptics wrong:
- Geoffrey Hinton: The British-Canadian cognitive computer scientist behind the renaissance of the neural network, popularizing the backpropagation algorithm that allows such networks to teach themselves by adjusting the weights of the various formulas after each mistake.
- Yann LeCun: A French computer scientist who pioneered the development of the so-called Convolutional Neural Networks (CNNs), which are the core technology for image recognition (recognizing objects in images and videos, recognizing faces, etc.) that powers software to run self-driving cars.
- Yoshua Bengio: A Canadian computer scientist and cognitive computer scientist whose work on the learning of sequential data and in particular natural language processing has enabled speech recognition and text modeling.
The Perfect Storm: Big Data and GPUs
Around 2010, the work of Hinton, LeCun, and Bengio collided with two massive global trends: the explosion of the Internet (big data) and the development of high-powered Graphics Processing Units (GPUs).
In order to learn from the massive amounts of data required to recognize patterns, a neural network needs a huge amount of data (e.g. images, text documents). And to process this data in a meaningful way, it requires a huge amount of mathematical calculations. The internet, with all its billions of images and text documents, became a treasure trove of data, and the computer graphics cards (GPUs) originally designed to power video games turned out to be the perfect hardware to run these massive calculations required by neural networks at tremendous speed.
The definitive turning point was the 2012 ImageNet contest, where a neural network created by the famous team of Geoffrey Hinton (founder of deep learning), Alex Krizhevsky, Ilya Sutskever, and Christian Farley called AlexNet crushed all the traditional, hand-coded computer vision algorithms by a massive margin. It was clear that deep learning had won.
7. Major Milestones in Artificial Intelligence History
To get a better feeling for the tremendous development that is taking place in AI right now, we are listing the most important milestones in the history of the field of AI:
8. From History to Strategy: Empower Your Business in the AI Age
Understanding who created artificial intelligence is not just about reading some history books. The companies of the future are going to be empowered by the same breakthroughs in business processes that were developed by the pioneers of AI.
AI has come a long way since its inception as a rigid set of manual rules, and today it exists in a data-driven, ever-changing form. The corporations that will rule the next decade are those who incorporate these historical innovations into their work on a daily basis.
While building an enterprise AI infrastructure from scratch is feasible for some, for most it is a daunting task and can be a huge time sink.
Elevate Your Operations with Elite AI Integration
That is where [Your Brand Name Here] comes in. We bridge the gap between the groundbreaking ideas from the founders of AI—Alan Turing and John McCarthy—and the current challenges of business growth of companies today.
Whether you are looking to create custom conversational interfaces, automatically process and analyze large datasets of information, or integrate generative AI into your current suite of enterprise applications with secure pipelines of functionality, we have a team of world-class AI engineers that can design and deploy custom applications that grow with your business.
Stop reacting to the future. Start building it. Contact [Your Brand Name]. Here today for a customized consultation and to help unlock your company’s true algorithmic potential.
Related Links:
Anthropic AI Tool
Advantages and Disadvantages of AI
Top 50 AI Tools Lists
AI Engineer Roadmap
Do visit our channel to learn More: SevenMentor
Author:-
Rohit Desai
Rohit Desai
Expert trainer and consultant at SevenMentor with years of industry experience. Passionate about sharing knowledge and empowering the next generation of tech leaders.