Understanding the Artifical. On the Future Shape of Artificial Intelligence

§The debate on the feasibility and applications of artificial intelligence has concentrated on the arguments about the concept of 'intelligence', largely neglecting the concept of the 'artificial'. Most of the recent literature is concerned with whether or not the concept of intelligence is transferable to the computer. Apart from the book by H.A. Simon, The Sciences of the Artificial (1969), no serious attempt has been made to define a conceptual framework for understanding the true nature of intelligent machines independently of the 'human' features claimed, or denied, for them. Understanding the Artificial aims to promote a 'second wave' of debate on AI which may help human culture to adapt itself to the artificial rather than to assimilate it according to a user-oriented philosophy. This collection of papers, edited by Professor Massimo Negrotti, discusses the potential losses and gains from AI if we abandon the original anthropomorphic pretension. §The authors aim to clarify the main differences between human knowledge processes and computer information processing and suggest the most plausible alternatives to the original ambitions of AI. The authors aim to redefine the epistemology and the objectives of the discipline by highlighting potential problems and possible solutions which should be accepted as typical of the concept of the 'artificial' rather than of an information-based imitation of man. In recent years a vast literature has been produced on the feasibility of Artificial Intelligence (AI). The topic most frequently discussed is the concept of intelligence, with efforts to demonstrate that it is or is not transferable to the computer. Only rarely has attention been focused on the concept of the artificial per se in order to clarify what kind, depth and scope of performance (including intelligence) it could support. Apart from the classic book by H.A. Simon, The Sciences of the Artificial, published in 1969, no serious attempt has been made to define a conceptual frame for understanding the intimate nature of intelligent machines independently of its claimed or denied human-like features. The general aim of this book is to discuss, from different points of view, what we are losing and what we are gaining from the artificial, particularly from AI, when we abandon the original anthropomorphic pretension. There is necessarily a need for analysis of the history of AI and the limits of its plausibility in reproducing the human mind. In addition, the papers presented here aim at redefining the epistemology and the possible targets of the AI discipline, raising problems, and proposing solutions, which should be understood as typical of the artificial rather than of an information-based conception of man.