Definition
GOFAI (Good Old-Fashioned AI), a term coined by philosopher John Haugeland, refers to the symbolic, logic-based approach to artificial intelligence that dominated the field from the 1950s to the late 1980s. It is based on the “Physical Symbol System Hypothesis,” which posits that intelligence consists of the high-level manipulation of discrete symbols according to formal logical rules.
Why It Matters
GOFAI was the first great attempt to build artificial intelligence, and its ‘brittleness’ in the face of real-world messiness remains a critical lesson; it teaches us that while logic is powerful, human-level intelligence likely requires the flexible, pattern-based learning found in modern neural networks.
Core Concepts
- Symbolic Logic: Representing knowledge as a set of axioms and using inference engines (like Boolean logic) to deduce new truths.
- Top-Down Intelligence: The assumption that intelligence can be pre-programmed by humans who understand the rules of a domain.
- Expert Systems: The most successful implementation of GOFAI, where human expertise was codified into “If-Then” rule bases (Expert Systems (AI)).
- Brittleness: The primary failure mode of GOFAI. Systems were unable to handle ambiguity, noise, or “common sense” situations that weren’t explicitly coded.
- Combinatorial Explosion: The problem where the number of possible logical combinations grows exponentially, making exhaustive search computationally impossible (Combinatorial Explosion).