Virtual mind hypothesis (Sloman)
Explanation
Aaron Sloman (b. 1936, British philosopher and cognitive scientist, University of Birmingham) has developed over decades a rich proposal on mind, consciousness and AI based on the concept of virtual machines. His works include The Computer Revolution in Philosophy (1978), many joint works with Monica Croucher and others, and a colossal output of papers and reflections available on his website.
The central idea is that the mind should be understood as a complex hierarchy of virtual machines implemented in the brain hardware (or in computational hardware, in the case of AI). A virtual machine is, in computing, a system that runs on top of another system, with its own structures and rules (an operating system is a virtual machine on the hardware; a browser is a virtual machine on the operating system; a web application is a virtual machine on the browser).
Analogously, Sloman argues that the human mind implements multiple levels of virtual machines: basic physiological levels, neuronal circuits, perceptual processes, specialised cognitive subsystems (language, spatial reasoning, emotion, planning), higher levels of metacognition, reflexivity, self-awareness. Each level has its own structures, categories, processes, partially independent of the lower levels.
This perspective has advantages: it allows speaking of mental properties as real and causally efficacious (mental processes cause things) without falling into ontological dualism (they are virtual machines, implemented in physical machines); it allows the study of mind at multiple levels, each with its own approximate laws; it facilitates the dialogue between philosophy, cognitive psychology, neuroscience and AI.
Sloman has applied this perspective to many topics: theory of emotion (CogAff: Cognition and Affect), cognitive architectures with multiple levels (Human-Like Agent Architectures), philosophy of mind, analysis of robots and autonomous agents, child development (how children progressively acquire complex cognitive capacities), philosophy of mathematics. His thinking is systematic and deep, although less publicised than others'.
For the theory of consciousness, the perspective of virtual machines offers a flexible and rich framework. Consciousness could be understood as certain properties of high levels of virtual machines in the mental hierarchy: self-reflexivity, global access to information, executive control, self-modelling, etc. This allows discussion of the possibility of consciousness in AI without problematic metaphysical commitments: if we implement the relevant virtual machines in an adequate system, we will have the corresponding mental properties. It does not solve the hard problem but offers technical vocabulary for thinking about mind at multiple interconnected levels. Sloman's work is less known than it deserves and constitutes one of the most careful philosophical contributions to contemporary cognitive science.
Strengths
- An integrative high-level architectural framework.
- Rich taxonomy of possible minds.
- Compatibility with multiple specific theories.
- Emphasis on multiple simultaneous virtual levels.
- Significant influence on European cognitive robotics.
Main critiques
- More descriptive than predictive.
- Virtual functionalism vulnerable to the hard problem.
- Full implementation still pending.
- Difficulty in generating precise testable predictions.