The world as we know it has come a long way in the last 100 years. Never before in the past has civilization achieved so much technologically in such a short period of time. Every day we get smartphones, computers, and applications that basically do all of our work for us. But all of that had a beginning, in a time that the technology of today would possibly be seen as magic.
In 1935 a British logician by the name of Alan Turing theorized about an advanced computing machine that has limitless memory. Combined with a scanner that would analyze this memory, it would create a memory of its own. This stored-program concept had implied the creation of a machine that would improve itself. The concept is known today as the Turing machine that all computers are basically based on. Even though he wrote a paper “Intelligent Machinery” in 1948, he never published it and later others used it for their work on AI.
A Chess Reference
While staying at Bletchley Park during World War II, Turing wanted to explain his theory of machine intelligence by using a chess reference. His theory was based on a premise that a computer who could play chess would search all the combinations of all the remaining moves. However, Turing only experimented with chess programs, but couldn’t really test them because he was lacking the main component – a computer.
In 1950, Turing introduced what is now commonly known as the Turing test. The test included 3 things: a human interrogator, a human foil, and a computer. All the communication between the participants was done via a display screen and a keyboard. The goal of the test is for the interrogator to guess which of the remaining two participants is a computer and which is human. It was achieved by asking different questions. The interrogator can ask any kind of question, and the computer is allowed to do anything to win – even lie. If the interrogator is unable to guess right, the computer is considered intelligent.
Turing could have set the ground for AI but there are many other scientists and theorists who have greatly contributed to its progress and development.
First AI Programs
One of the first AI programs that ended up working successfully was created in 1951. It was created by Christopher Strachey at the University of Manchester. The program could play a game of checkers and it was run on Ferranti Mark I.
Problem-solving and logical reasoning has always been an imperative part of artificial intelligence. In this regard, a breakthrough was made in 1956 with a program called a Logic Theorist. It was designed to prove theorems from Principia Mathematica.
MIT’s Joseph Weizenbaum and a psychiatrist Kenneth Colby from Stanford University, made two AI programs that were the first ones to have an intellectual dialogue. The first program was named Eliza and it had the role of therapist, while the second one named Perry was a paranoiac. Even though some psychiatrists when communicating with Perry couldn’t decide if it was a real person or a machine, neither Perry nor Eliza couldn’t be considered intelligent.