In the early 1970s, research on artificial intelligence to a large extent had rejected its
original goal of finding generic and universal rules that produced intelligence. After
some early successes in the 1950s and 1960s, researchers became convinced that
intelligence required domain specific knowledge. This, in turn, required new
approaches in representing knowledge in forms that could be processed by a
computer. As a result, in the 1970s the focus in AI research moved to systems that
followed a relatively simple logic but which had detailed knowledge of their domain
of application (McCorduck, 1979).