History of AI: the key milestones

Artificial Intelligence (AI) comprises a set of sciences, theories and techniques that, as its main objective, aims to imitate the cognitive capacities of human beings. The history of AI begins at the start of the Second World War and its developments are intimately linked to those of informatics, which have led computers to execute increasingly complex tasks which could first only be undertaken by humans.

In this article, we’re going to go through the history of Artificial Intelligence so that you can learn about its evolution and future expectations, especially for those wanting to get training in this fascinating discipline. Want to join us for the ride?

 

The beginning of AI history until the 60s decade: the discipline is born

The period between 1940 and 1960 was decisively shaped by the sped-up technological developments driven by the Second World War and the desire to understand how machines could be used to ease human work.

For Norbert Wiener, a robotics pioneer, the goal was to unify mathematical theory, electronics and automation in a unified theory. Just before him, Warren McCulloch and Walter Pitts had developed the first mathematical and computer model for neurons in 1943. This worked as one of the referential basis for ensuing developments within Artificial Intelligence such as deep neural networks (currently on the rise).

Another initial milestone in the history of AI can be found in an article written by Turing in 1950: Computing Machinery and Intelligence. In this article, the British scientist described an imitation game where a human was challenged to distinguish whether his dialogue on a teletype was produced with a human or a machine. Besides, and along with John Von Neumann, Turing was the founding father of the technology behind the concept of Artificial Intelligence, providing a formal framework for the architecture of contemporary computers and demonstrating this was a universal machine, able to execute any program.

Alan Turing (1912 – 1954) committed suicide two years after having been subjected to chemical castration for being homosexual. In 2016, the British Government apologized and exculpated him. 

Although Turing, along with John Von Neumann, was very influential in many of the ensuing developments, the conceptualization of Artificial Intelligence can be mainly attributed to John McCarthy, which defines it as “the construction of computer programs dedicated to tasks currently executed in a more satisfactory manner by human beings as they require of high-level mental processes such as: perceptive learning, memory organization and critical reasoning.”

 

The history of AI in the 80s decade

Another key development in the history of Artificial Intelligence took place in the 70s with the advent of the first microprocessors. This door was first opened in 1965 by Dendral, and then in 1972 by MYCIN. These two systems were based on an inference engine, which was programmed to become a logical mirror to human reasoning. After entering data on it, the engine provided answers that showed a high level of experience.

This triggered the promise of massive development for AI, but it wasn’t until the end of the 80s decade when a small-scale boom took place within AI systems programming. However, maintenance and development were still too costly in economic terms, and the results weren’t as expected.

ChipTest, the Deep Blue predecessor, was developed in 1985 as a project by Carnegie Mellon University at Pittsburgh, Pennsylvania. Not much longer after, in 1989, this evolved into Deep Thought, more powerful and efficient. 

 

The late 90s: Deep Blue

In May 1997, Deep Blue, an AI system developed by IBM, won a chess game against champion player Garry Kasparov. Deep Blue was based on a systemic algorithm using brute force, where all potential movements were evaluated and pondered.

The first games between Kasparov and Deep Blue were played in 1996, the final winner being the one that won the most of a total of 6 games. This time, the human won. A year later, when IBM improved their AI system, the computer was able to beat the chess champion.

Kasparov’s defeat is one of the key feats in the history of Artificial Intelligence, its possibilities and its later development from the first decade in the 21st century. Although the truth is that Deep Blue had only achieved to take in a very specific feature (the rules of chess) so that it was all very distant from being able to model the complexity of the real world.

 

Big Data: the boom in the history of AI

Two factors explain the new boom experienced by Artificial Intelligence, mainly from the year 2010: 

• Access to big volumes of data: before, image-classifying algorithms were only able to, for example, recognise cats if manual samplings were first done. Today, a simple Google search is able to find millions of these images.

• High efficiency in graphic card processors. This allows for the acceleration of learning algorithm’ calculations. Tasks that used to take weeks have been cut down to days, hours or minutes thanks to the processing abilities of modern systems.

How did these two wonders originate? A complete shift of paradigm for systems. The approach has become inductive: now it’s not just about codifying rules, but also letting machines discover them independently by correlation and classification, on the basis of a massive quantity of data: Big Data.

Among the key techniques in automatic learning, deep learning seems to be the most promising for a wide variety of applications.

Suddenly, almost overnight, a majority of research teams resorted to this technology which, today, involves a huge number of advantages. This type of learning has also allowed for considerable advancements in text recognition, although there’s still a long way to go to generate systems that are able to understand texts.

 

Get trained in Artificial Intelligence!

Now you know a bit of the history of AI and this discipline’s background. As you might have learned, this field’s history goes back in time longer than people tend to think, all while new technologies are opening a wide variety of options that are still to be discovered for the future. A fascinating discipline that requires quality training by enrolling in the best educational institutions and getting instructed by the most proficient professionals.

Learn more about our offer at Tokyo School and get ready to become a much-demanded professional for your future with our Artificial Intelligence specialization course. Fill in the form below and request more information now. We can’t wait to meet you!