Granola Cruncher Flashcards
Behaviorism
Discusses Aristotle and the practical syllogism, emphasizing the role of
belief and desire in action production. It highlights that behaviourism, dominant from
~1900 to 1950, focused on the interplay between belief and desire leading to actions
Alan Turing and Computer Theory
Explores the parallels between brain function and
serial computers, and the origins of cognitive science. It delves into functionalism and
the possibility of implementing algorithms to replicate human thoughts and actions.
Chomskian Linguistics
Addresses Noam Chomsky’s idea of an innate universal
human grammar and the distinction between syntax and semantics, noting that
computers excel at syntax but struggle with semantics.
Multiple Realizability and Jerry Fodor’s Views
Describes Fodor’s computationalist
perspective, discussing propositional attitudes, the computational nature of
psychological processes, and the dependency of psychological laws on the content of
beliefs and desires.
Connectionism
Examines the classical viewpoint on thinking, emphasizing symbolic,
serial processes and logical rules. It discusses the concept of mental representations
being sentence-like and word-like
Problems with Serial Computations
Highlights the limitations of serial computations,
including their monotonic nature and inability to revise assumptions leading to
contradictions.
Frame Problem:
Discusses the logical and epistemic challenges in determining how
changes in one aspect (like color) affect other aspects (like position) and the brain’s
difficulty in resolving these ambiguities.
Connectionism (Part 2)
Focuses on neural networks and their resemblance to brain
function, discussing the integration of syntax and semantics in connectionist networks,
which are seen as distributed parallel processing networks.
Questioning Human Thinking and AI Models:
Questions the suitability of
connectionist networks for modelling human cognition, suggesting that they might
represent a form of thinking but not necessarily human-like thinking.
Rodney Brooks’ Perspective on Cognition
Suggests that traditional views on
cognition, particularly the reliance on internal symbolic representation, might be
misguided. Brooks advocates for a model of cognition centred on navigating the world
without internal symbolic representation.
Dynamical Systems Theory
Explains this theory as modelling phenomena through
differential equations, focusing on adaptation to stable yet changing environments. It
uses the Watt governor as an example to illustrate real-time behaviour adjustment and
the continuous, component-free nature of thought.
Chapter 1 of Jerry Fodor’s “The Elm and the Expert: If Psychological Processes
are Computational, How Can Psychological Laws Be Intentional”:
Compares old folk psychology to new Alan Turing psychology. Has three theses’:
● There are intentional laws
● Semantic properties are informational
● Intentional laws are computationally implemented
The Connectionist and Serial Computationist approaches represent two different
models of information processing:
Connectionist Approach:
● Based on neural networks mimicking brain structure.
● Uses parallel processing, handling multiple operations simultaneously.
● Adaptative and good at pattern recognition, like image and speech
recognition.
● Learns from experience by adjusting connections between nodes.
Serial Computationist Approach:
● Involves sequential, step-by-step information processing.
● Focuses on symbol manipulation and following explicit rules.
● Suitable for tasks requiring precise, logical outcomes.
● Struggles with complex, unstructured tasks.
In summary, connectionism excels in tasks requiring adaptability and pattern
recognition, resembling brain-like processing, while serial computation is more aligned
with traditional, rule-based computing tasks.
