ABSTRACT
This chapter explains what Intelligence means and how we should think about it. It looks at the development of intelligent programs called 'expert systems'. The chapter starts with a simple description of expert systems and then sets up a model of human culture intended to bridge the gap between the way the 'knowledge engineers' think and the way sociologists/philosophers of knowledge think. Using this model, a four-stage classification of expert systems, with predictions of the likely success of the different types is developed. The chapter concludes by setting the discussion in the context of the larger debate about the possibility of AI, and shows how intelligent machines may come to seem to think like us, and how this illusion might be dispelled. The crucial ideas are the inescapable input of an end user in an intelligent interaction, and the notion of 'behavioural co-ordinates of skill and social action'.
