Blog

Introduction to Artificial Intelligence #1: the origins

Artificial Intelligence (AI) raises many questions. And for a good reason: it is an extremely broad area for which there are many definitions and which disrupts many aspects of society.

If we hear a lot about AI today, it’s because it has an increasingly more significant place in our lives: indeed, we use tools based on AI every day. For instance, it helps us with automatic word completion on mobile phones, it allows the recommendation of press articles or information in a newsfeed or it is used to optimize the price of airplane tickets.

In this first article of our series on Artificial Intelligence, we will begin with its history: where does it come from and what are its origins?

Origins

In 1955, John McCarthy is the first to use this term and to define it as follows: “For the present purpose the artificial intelligence problem is taken to be that of making a machine behave in ways that would be called intelligent if a human were so behaving”[1].

This definition echoes the hope that scientists had at the time to see an artificial human brain created, as it was indeed the case a few decades later. It also highlights the fact that according to the threshold defined to discriminate what is intelligent and what is not, the term AI will cover different subjects.

AI is not a new concept: even before 1955, it can be traced back to several centuries when the first automata were created. Originally, these automata were trying to reproduce some animal behaviors; in 1515, Leonardo da Vinci designed a lion able to walk and turn its head as if it was roaring. Of course, we now view this behavior as extremely simple and we are not likely to accept the fact that it may be an “intelligent” behavior. Yet, it was just an object made of wood and metal which walked![2]

Evolutions and perception of AI over time

It took more than a century to observe a significant advancement of AI with the appearance of the calculator, invented by Blaise Pascal in 1645. He writes on this subject in his Thoughts: “The arithmetic machine’s effects are more similar to thought than what all animals do; but it does not do anything to make us say it has will, like animals[3]“. Today’s AI has not evolved: it does not have its own will and it only performs what it is programmed for. In fact, this should not exempt us from paying attention to the ethical issues raised by this tool.

This is one of the reasons why the phrase “Artificial Intelligence” is sometimes criticized by some of the players in the field, like Dr. Kate Crawford. She highlights the fact that although the tools are higly technical, they are designed by technicians to perform one task only: they have no consciousness and no will. They are extremely specialized which means they would be incapable of performing varied tasks!

The automaton of Vinci brings us back to this question: what can be considered intelligent? In 1950, Alan Turing, known for decrypting the Enigma[4]machine during World War II and considered one of the fathers of computers as we know them today, invented the Turing test[5]. The principle is to make someone speak with two other entities (in writing, to avoid vocal synthesis biases): one is a computer, the other is a human. If the first person is not able to determine who the human is, with no time limit, the AI passes the test. We are far from automata…

A mixed reception?

However, after its official birth in the 1950s, the progress achieved is inconstant: AI has followed a chaotic path, sometimes perceived as a rich and promising domain, sometimes illusory and utopian. This lasts until the 2000s, when AI finally meets the success it has today, both technical and social.

In the next articles of our series, we will see what factors changed the game in the 2000s. We will also dive into the technical world of AI to better understand why it is difficult to define it and why there are many different use cases.

[1]Source: A proposal for the Dartmouth summer research project on artificial intelligence

[2]One can also think of the Greek myth of Galatea, in which Aphrodite gives life to a statue: can one have signs of humanity from inanimate objects? It would be excessive, however, to choose this myth as the origin of artificial intelligence.

[3] “Pensées, fragments et lettres de Blaise Pascal”

[4]Enigma is a cryptographic machine used by Nazis in Germany during World War II 

[5]Also called the game of imitation, it is described by Turing in this article