Computers as Metaphor, Minds as Computers: Notes Towards a Dysfunctional Robotics
Symposium:
Session Title:
- Signs of Life: Human-Robot Intersubjectivities
Presentation Title:
- Computers as Metaphor, Minds as Computers: Notes Towards a Dysfunctional Robotics
Presenter(s):
Venue(s):
Abstract:
Panel: Signs of Life: Human-Robot Intersubjectivities
This paper will present a range of ideas underpinning the development of John Tonkin’s new project, a series of dysfunctional robots that explore different approaches to thinking about cognition and perception. Computational theories of mind have been used both by cognitive scientists as model of how to build an electronic mind, and by cognitive psychologists as a means of understanding the human mind. They see the mind as an information-processing system and thought as a form of computation. These symbolic approaches to thinking about the mind have been challenged by more embodied and embedded approaches to cognition and perception. This has been reflected through the development of a number of bottom-up approaches to AI and robotics, such as neural networks and behaviour based robots that are based on ideas of reactivity and situatedness rather than higher level symbolic modelling. The nervous robots that are being built for this project awkwardly hybridise bottom-up AI approaches with more classical symbolic approaches that use high level symbols drawn from a folk psychology conception of the mind as being the home of internal mental processes such as motives, desires, phobias and neuroses. They use a range of computational approaches, for example Brooks’ subsumption architecture, to create layered hierarchies of stimulus / response reflexes. Examples include a claustrophobot and an agoraphobot, as well as needy/dismissive robots based around attachment theory. One of the aims of this project is to explore the lower boundary of computational complexity that still evokes some sort of self-identification and response in the audience.