Cognition: Brains, Computers, and Feedback Metaphors

These flashcards cover various metaphors for the brain, historical frameworks of cognitive science (such as behaviorism and cognitivism), and key concepts including representation, feedback loops, situatedness, and embodiment.

Computationalist Cognitivism

The predominant modern metaphor in the scientific community and general public that the brain functions similarly to a digital computer that processes information.

Dualism

The idea that cognition arises from the interaction of two distinct aspects of reality: physical (material systems) and mental (the soul).

Behaviorism

The movement focused on understanding cognition by observing how agents interact with their environment, emphasizing observable behaviors over internal introspection.

Introspection

The idea of figuring out how one is thinking by examining one's own experiences, which behaviorists rejected because memories and thoughts can be misleading or different from actual experiences.

Cybernetics

A movement emerging in the 1940s focused on feedback systems and how creatures accomplish adaptive behavior to satisfy goals.

Cognitivism

The view that cognition involves the manipulation of internal mental representations and processes.

Connectionism

A framework for understanding cognition that focuses on neural network-like structures.

Inaction

An approach that attempts to explain how a system can have its own needs and goals and act to satisfy them based on its interaction with the environment.

Representation

Internal symbols or models of the world, such as the position of pieces on a chessboard, used by algorithms to determine action.

SLAM

An acronym for simultaneous localization and mapping, where a robot builds an internal model of its environment while simultaneously tracking its location within it.

Shurtleu

A 1960s simulation program that could communicate via natural language to manipulate cubes and pyramids within a constrained box.

Micro worlds

Very constrained simulated scenarios used to develop AI, where engineers identify all possible interactions and important object properties at the time of programming.

The Representation Problem

The challenge in classic AI where human engineers, rather than the AI, must decide which parts of the environment are important to represent and solve a problem.

The Linear Model

A sequential mode of operation involving a series of decoupled phases: sense, model, plan, and act.

The Sandwich Model of Cognition

A view that intelligence is located primarily in the middle layer of internal modeling and planning, between the sensors and the motors.

Sensory motor feedback

An ongoing loop where an action affects the relationship with the environment, which in turn affects the sensors and future responses.

Braitenberg vehicles

Simple robots with direct wires connecting sensors to motors that accomplish tasks like moving toward light without a brain or internal model.

Optic flow

The pattern of how visual bits move from the center of vision toward the periphery as an agent moves forward, used for navigation without a map.

Guitar feedback

A metaphor for cognition where a signal (sound) excites itself through a cyclic interaction between a guitar, amplifier, and the air, creating tone through environmental manipulation.

Situatedness

The idea that the world is its own best model, and raw sensory data can be used to solve problems directly without building complex internal simulations.

Embodiment

The perspective that the physical body plays an essential role in cognition, which was often ignored in early disembodied AI like chess programs.

Passive dynamic walker

A robot with no brain, motors, or sensors that can walk down a hill using only hinges, its own physical structure, and gravity.

Proprioceptive sensors

Internal sensors, such as a compass, that detect an agent's own movement or orientation.

John Searle

The author who noted in the 1980s that humans use the latest technology — like telegraphs, hydraulic systems, or computers — as a metaphor for the brain.