Contrary to generally accepted thought, intelligent agents are not a recent innovation. In fact, they are an old phenomenon that comes into fashion at regular intervals.
The ancestor of all agents is named Eliza, who has her origin well before the Internet explosion. Eliza was born in 1966 at Massachusetts Institute of Technology (MIT), thanks to the work of a German pioneer: Professor Joseph Weizenbaum. She was created in order to simulate a conversation with a psychotherapist. With only 240 lines of code, she is able to answer many requests. Her operation is very simple: this agent analyzes each question that is asked, and quickly identifies the keywords that she knows. With each recognized keyword, there are one or more standard corresponding answers. In spite of her age and thanks above all to the arrival of the Web, Eliza has been "maintained" in life: many Internet sites make it possible to have access to her "advice" from just the 240 original lines of codes!
However, after a few minutes of discussion, you will quickly discover the limits of this agent. Eliza is part of the "ChatterBot" family and her descendants have acquired considerable technological maturity, being skillfully able to keep your attention in a conversation.
During the Seventies, the evolution of intelligent agents was naturally related to progress in artificial intelligence (AI). After the initial tentative steps, the first models began to impose themselves. In 1977, Carl Hewitt proposed the concept of an object with autonomous behavior, able to answer requests from other objects.
As the goal of this work is primarily pragmatic, we will develop neither the theoretical aspects, nor the technical aspects with respect to the operation of the agents. For those who wish to investigate these subjects, there are many free Internet sites that can be consulted.