The A.I. was nominated like this for the first time in 1956 to name a meeting of a group of researchers from Dartmouth, a US Ivy League college. At that very first meeting, the participants discussed how computers will soon perform all human activities that require intelligence, including playing chess or other games, composing good music, and translating texts from one language into another. These pioneers were very optimistic, though their aspirations were not surprising. The interest in build intelligent machinery has been a longstanding concern for man, both in literature and computing. The first computers of the 1940s were described as thought machines or electronic brains. title
The scientist British Alan Turing, the computers father – scientifically speaking- from the beginning had no doubt that one day the computers would start to think. His famous 1950s article introduced the Turing test, in which a man has to discern between another person and a car without seeing them, only on the basis of their answers to a series of questions. All Research in the field of Artificial Intelligence from the 1950s to the 1970s focused on writing programs to perform tasks that required and used human intelligence. An example was the computer game of American Arthur Samuels, the Checkers game. The program got better by analyzing winning positions and quickly began to defeat Samuels. However, what went for the checkers did not use for more complicated games like chess. In mathematics the Artificial Intelligence had a spectacular evolution, a research project addressed mathematical analysis introductory problems, especially symbolic integration. In a very short period of time the symbolic integration has become a problem solved and the programs for solving it were no longer interpreted as Artificial Intelligence. In contrast to the checkers or integrals, translation programs or voice recognition have made little progress during the time. It is very difficult to use the power of computer processing efficiently and reach the perfection in this field.
Interest in artificial intelligence increased spectacularly in the 1980s, with expert systems. Successes were achieved with programs that performed medical diagnosis, geological mapping for minerals, and configuration of computer requests. Although useful for a small set of problems, expert systems were neither robust nor general and required detailed expert knowledge to be created. The programs did not show ordinary intelligence and so, after a spectacular increase of interest in artificial intelligence, there was a decline in the 1990s. In the last years with the increase in computer processing power, voice recognition and language processing have improved considerably. New algorithms have been created that focus on statistical modeling techniques rather than simulation of human processes. Progress continued with the creation of voice-controlled assistants such as the Apple Siri program and the Google Ok program. Nobody thinks that a computer really understands a language at the present, despite the development of programs like conversation robots. There are definite limits on what Siri and Google Ok can process, and translations do not always highlight the context. Facial recognition was also considered a challenge in the 70’s. The programs were hopeless at that time. Today, however, many Socials can identify people from numerous “tags” and dedicated software can recognize the faces captured by camcorders. However, it is rather advanced statistical methods than intelligence, but is considered anyway Artificial Intelligence.