The Historical Development

If you follow the media coverage of artificial intelligence, you get the impression that the development of this discipline has only recently begun with great vigor. But that is completely wrong. Like any other discipline, the development of artificial intelligence can look back on a long and interesting history. It begins with G. F. Leibniz (1646–1716). He replaced the decimal system with the dual system without having an application in mind. Addition led him to an operation known in logic as antivalence (either …or). The next fundamental step was the construction of propositional logic by G. Boole (1815–1864). His intention was to make the truth or falsity of sets of propositions calculable. He started from the fact that repeating a statement does not change its truth or falsity. He expressed this fact using the equation a \(\times \) a = a. This equation only has the two solutions a = 0 and a = 1, whereby it is customary to use the value 1 for the truth and the value a = 0 for the falsity of a statement. This “Boolean algebra” is used today in two ways, firstly as logic in the original sense and secondly as the basis for the design of microelectronic circuits. The term “artificial intelligence” was used explicitly for the first time in the invitation to the 1955 Dartmouth Conference. Particularly noteworthy are the topics of “neural networks,” the “complexity of computation,” “self-improvement,” and “creativity.” The increasing development of the computer world was accompanied by ever more powerful programming languages. The most significant contribution was made by the English mathematician Alan Turin (1912–1954). He created the first precise definition of an algorithm with the concept of the Turing machine. At the same time, performance increased enormously. According to Moore’s law, the performance of hardware doubles in about 1.5 years.