The Definitive History of AI: Part 4 - The Boom of Expert Systems (1980-1987)
Welcome to the fourth installment of our seven-part series on the history of Artificial Intelligence. Today, we'll explore the period from 1980 to 1987, known as the "Boom of Expert Systems" - a time when AI saw a resurgence in enthusiasm and funding, primarily driven by the success of rule-based expert systems.
The Rise of Expert Systems
Expert systems emerged as a promising subfield of AI during the 1970s and gained significant traction in the 1980s. These systems were designed to emulate the decision-making ability of a human expert in a specific domain. Unlike earlier AI systems that aimed for general intelligence, expert systems focused on narrow, specialized areas of knowledge.
Key characteristics of expert systems include:
- A knowledge base containing domain-specific information
- An inference engine to apply the knowledge to specific problems
- A user interface for interaction with non-expert users
Notable Expert Systems
Several expert systems developed during this period demonstrated the practical potential of AI:
- XCON (1980): Developed by Digital Equipment Corporation (DEC), XCON was used to configure VAX computer systems. It was one of the first commercially successful expert systems.
- PROSPECTOR (1979-1984): This system aided geologists in mineral exploration and successfully predicted the location of a molybdenum deposit worth millions of dollars.
- DENDRAL (1965-1983): Although its development began earlier, DENDRAL continued to evolve during this period. It was used to identify unknown organic molecules based on mass spectrometry data.
- CADUCEUS (1982): An expansion of the INTERNIST-I system, CADUCEUS was capable of diagnosing complex internal medicine cases.
- DELTA/CATS-1 (1984): Developed by General Electric, this system was used for maintenance of diesel-electric locomotives.
The Japanese Fifth Generation Computer Project
In 1982, Japan launched an ambitious 10-year plan known as the Fifth Generation Computer Project. The goal was to develop computers that could perform AI tasks like natural language processing, speech recognition, and machine learning. This project, with its substantial government funding, sparked a wave of AI enthusiasm and investment around the world.
While the project ultimately fell short of its lofty goals, it played a crucial role in revitalizing AI research globally. It prompted other countries, particularly the United States and United Kingdom, to increase their own investments in AI to avoid falling behind.
Advances in AI Technologies
The 1980s saw significant advances in various AI technologies:
- Knowledge Representation: Techniques like frames and semantic networks were developed to represent complex knowledge structures.
- Natural Language Processing: Progress was made in parsing, understanding, and generating human language, although significant challenges remained.
- Machine Learning: While not as prominent as expert systems, machine learning techniques like decision tree learning began to gain attention.
- Neural Networks: Interest in neural networks was rekindled, setting the stage for their resurgence in the following decades.
AI Programming Languages and Tools
The boom in expert systems led to the development of specialized AI programming languages and tools:
- LISP Machines: Specialized computers optimized for the LISP programming language, popular in AI research.
- Prolog: This logic programming language gained popularity for AI applications, particularly in Europe and Japan.
- Expert System Shells: Tools like OPS5 and CLIPS made it easier to develop expert systems without starting from scratch.
Commercialization of AI
The success of expert systems led to increased commercialization of AI:
- Companies like IntelliCorp, Teknowledge, and Symbolics were founded to develop and sell AI technologies.
- Established tech companies like IBM and DEC invested heavily in AI research and development.
- The AI market grew from a few million dollars in 1980 to billions by 1988.
Challenges and Limitations
Despite the enthusiasm, expert systems and other AI technologies of this era faced significant challenges:
- Knowledge Acquisition Bottleneck: Extracting and codifying expert knowledge proved to be time-consuming and difficult.
- Brittleness: Expert systems often performed poorly when faced with problems slightly outside their specific domain.
- Scalability: As knowledge bases grew larger, maintaining consistency and performance became increasingly challenging.
- Limited Learning Capability: Most expert systems couldn't learn from experience or update their own knowledge bases.
Cultural Impact
The AI boom of the 1980s had a significant impact on popular culture:
- Movies like "Blade Runner" (1982) and "The Terminator" (1984) explored themes of artificial intelligence and its potential consequences.
- William Gibson's novel "Neuromancer" (1984) popularized the concept of cyberspace and artificial intelligences existing in digital realms.
- AI concepts began to appear more frequently in mainstream media, both as subjects of fascination and sources of concern.
Conclusion
The period from 1980 to 1987 marked a significant resurgence in AI research and development, primarily driven by the success of expert systems. This era demonstrated the potential of AI in solving real-world problems and attracted substantial investment from both government and private sectors. However, it also revealed new challenges and limitations that would need to be addressed in the coming years. The boom of expert systems set the stage for further developments in AI, even as it foreshadowed some of the difficulties that would lead to the second AI winter in the late 1980s and early 1990s.