Aims and Objectives


Spring 2000 saw the Society for the Study of Artificial Intelligence and the Simulation of Behaviour (SSAISB) hold their annual symposium at the University of Birmingham. One of the principal workshops ran under the title “How to Design a Functioning Mind”. Organised by Professor Aaron Sloman, the workshop attracted a set of international speakers. The over-riding theme was the adoption of a multi-disciplinary approach to the long term problem of designing a human-like mind, whether for the scientific purpose of understanding human minds or some engineering purpose. Its purpose was to bring together people interested in building bridges between various kinds of partial studies, with the long term goal of understanding, at least in principle, how to build a complete mind. This text has very similar objectives. It draws on those authors present at the workshop and others to provide an update on their perspectives for this collected text.

Objectives and Mission

The main objective of this collection is to present an overview of the research area of artificial minds at the start of the twenty first century. As such it will draw on current and prospective work from pivotal researchers in the area. It will include perspectives from philosophy, psychology, cognitive science and artificial intelligence. Rather than present a unified model of mind, it will invite a diverse collection that mirrors the academic discipline itself. The objectives and mission of the text are in effect a rephrasing of those associated with the original workshop.

The text presents a multi-disciplinary approach to the long term problem of designing a human-like mind, whether for the scientific purpose of understanding human minds or some engineering purpose.

Much research in AI is fragmented. People work on language, or vision, or planning, or learning or mathematical reasoning, without necessarily asking how their models could be combined with others in a fully functioning mind; or they discuss multi-agent systems where the agents have only very simple collections of capabilities.

Much research in psychology is equally fragmented: investigating particular capabilities and how they are affected by environmental factors, or brain damage, or gender, or age, etc.; for instance linguistic or visual or problem solving or memory or motor control capabilities.

Moreover such research often produces interesting empirical results without leading to a theory that is deep or precise enough to be the basis for a design for a working system.

Some philosophers also think about these topics and attempt to analyse the concepts involved in talking about minds, or necessary or sufficient conditions for various kinds of mentality, but without doing so at a level that might guide an engineer attempting to design a mind. Some of them produce arguments claiming to show that the task is impossible, but without formulating the arguments in a manner that could convince a computer engineer.

The questions that are addressed by this text are:

·         What is mind?

·         What are theories of mind?

·         What are computationally plausible theories of mind?

·         What are computationally plausible designs based on these theories of mind?

·         What are computationally plausible architectures and systems to support theories of mind?

·         What are computationally implemented architectures and systems based on theories of mind?

·         What kind of tools can we use in producing and implementing such designs?

·         How do we know when we are successful in producing artificial minds?