Symbolic AI
An approach to artificial intelligence based on manipulating human-readable symbols and explicit rules to represent knowledge and solve problems.
Also known as: GOFAI, Good Old-Fashioned AI, Classical AI, Rule-Based AI
Category: AI
Tags: ai, history, computer-science, reasoning, knowledge-representation
Explanation
Symbolic AI, also known as Good Old-Fashioned AI (GOFAI), is an approach to artificial intelligence that represents knowledge using human-readable symbols (words, numbers, logical expressions) and manipulates them according to explicit rules. It was the dominant paradigm in AI from the field's founding in the 1950s through the 1980s.
**Core ideas:**
- **Physical Symbol System Hypothesis**: Proposed by Allen Newell and Herbert Simon in 1976, this states that a physical system has the necessary and sufficient means for general intelligent action if it can create, modify, and manipulate symbols. In other words, intelligence is fundamentally symbol manipulation
- **Knowledge representation**: The world is represented using formal structures such as logical propositions, semantic networks, frames, and ontologies
- **Inference**: New knowledge is derived from existing knowledge using logical rules (deduction, induction, abduction)
- **Search**: Problem-solving is modeled as searching through a space of possible states
**Key achievements:**
- **Logic Theorist (1956)**: One of the first AI programs, proved mathematical theorems
- **DENDRAL (1965)**: Inferred molecular structures from mass spectrometry data
- **MYCIN (1976)**: Diagnosed bacterial infections and recommended antibiotics
- **Expert systems (1980s)**: Commercial applications encoding domain expertise in rule-based systems
- **Cyc (1984-present)**: Ambitious project to encode common sense knowledge
**Strengths:**
- Transparent and explainable: you can trace exactly how conclusions are reached
- Good at representing structured knowledge and logical relationships
- Can work with small amounts of data when rules are well-defined
- Captures expert knowledge in a reusable, auditable form
**Limitations:**
- Brittle: fails on inputs outside its rules
- Knowledge bottleneck: extracting and encoding expert knowledge is extremely labor-intensive
- Struggles with uncertainty, ambiguity, and real-world messiness
- Does not learn from experience; requires manual rule updates
**Symbolic AI vs. connectionism:**
The tension between symbolic and connectionist (neural network) approaches has defined much of AI's history. Symbolic AI excels at logical reasoning and structured knowledge; connectionism excels at pattern recognition and learning from data. Modern AI increasingly combines both: neural networks handle perception and pattern recognition while symbolic methods handle structured reasoning and planning.
Related Concepts
← Back to all concepts