Michael Rescorla - Computational theory of mind

A theory positing that the mind functions similarly to a computational system, particularly a Turing machine.

Refers to the neurophysiological properties of an organism’s brain or body.

Describes mental states in terms of their representational properties.

Algorithm: A step-by-step procedure for solving a problem. Turing machine: An abstract model of a computing device with infinite time and storage, manipulating symbols. Memory Structure: Linear arrangement with infinite cells. Processor: Moves one cell at a time, accessing memory. States: Finite machine states defining processor behavior. Operations: Write, erase, move left or right. UTM (Universal Turing Machine): Can replicate any other Turing machine.

Mechanization of thought, examples include Deep Blue and driverless cars. AI for decision-making, such as Bayesian decision theory.

Views mental activity as Turing-style computation. Mental processes compared to Turing machines. Differentiates from physicalist and dualist views.

Mental states are understood as functional states within a system. Posits mental activity as implementing a probabilistic automaton.

Mentalese is compositional. Discusses productivity and systematicity in thought.

Comprise input, hidden, and output nodes. Learning involves adjusting weights. Includes feedforward and recurrent networks.

Classical computation involves rule-governed symbol manipulation, while neural networks are non-symbolic.

Representational mental states have content that represents the world. Discusses intentional realism, eliminativism, and interpretivism.

Uses biologically realistic neural networks to model cognitive phenomena.

Formal-syntactic conception of computation.

Mental content is partly determined by factors outside the thinker.

No rigid demarcation between computational and intentional description.

Information-processing, function evaluation, and structuralism.

Functional mechanisms and mechanistic explanation.