Cognitive Science and AI

AI and cognitive science co-evolved, each informing the other. This sub-article expands Section 2.1 of Philosophy.

Cognitive Architectures

• ACT-R and SOAR model human cognition computationally.
• These systems emphasize working memory, production rules, and learning.

Two Traditions

• Symbolism: thought is symbol manipulation; inspired by logic and language.
• Connectionism: thought is distributed neural activity; inspired by brains.

Dual-Process Theories

• Fast intuition vs. slow deliberation.
• Modern AI hybrids combine fast learned heuristics with deliberate reasoning loops.

The Reverse: AI as a Model of Mind

• Neural networks are used to test hypotheses about perception, memory, and language.
• LLMs are studied as possible models of aspects of human cognition.

Why It Matters

Cognitive science provides AI with theories of mind, and AI offers computational models to test those theories.

Sources

• Anderson, J. (1983). The Architecture of Cognition.
• Newell, A. (1990). Unified Theories of Cognition.

Back to Philosophy.