Mind Design – John Haugeland

This is an abstract of the first chapter of John Haugeland’s Mind Design II, an edited collection of important works in cognitive science. It provides an overview of the framework used by cognitive scientists and, in particular, those interested in designing intelligent artifacts.

Abstract

The goal of “mind design” is to understand thinking and intelligence in terms of the design of a mind. Intelligence makes sense only in perspective (no individual atom can be called intelligent, but systems as a whole may be); possible approaches to the question of intelligence are the Turing Test and “the intentional stance,” the rationality of a pattern of behavior betraying underlying belief or understanding. Traditional (GOFAI) approaches to building intelligent artifacts have attempted to create computers whose behavior (such as it may be) may be interpreted as intentional and rational, serving to reason through problems. One new approach, connectionism, uses non-formal pattern processing to tackle intelligence. Another new approach, embodied and embedded AI, posits that a system must be integrated with its environment in order to act intelligently. None of these approaches tackles directly issues like “true” intentionality and understanding, but they may address them nonetheless.