Which trends characterized AI in the 1980s?

Which trends characterized AI in the 1980s?

• A. Expert systems, case-based reasoning, and revival of neural networks with backpropagation
• B. Quantum computing breakthroughs
• C. Neural networks without backpropagation
• D. Symbolic AI dominance

Answer: (A)

Three defining trends defined AI in the 1980s: expert systems, case-based reasoning, and the revival of neural networks with backpropagation. Expert systems represented domain knowledge as explicit rules and used inference engines to apply that knowledge, which led to practical, commercial AI applications in areas like diagnostics and configuration. Case-based reasoning offered a different approach by solving new problems through reusing and adapting solutions from similar past cases, providing flexible problem-solving for ill-defined tasks. At the same time, neural networks experienced a revival because the backpropagation algorithm made training multi-layer networks feasible, enabling learning from data and enabling patterns to be discovered without explicit programming of rules. These three threads together capture the decade’s distinctive AI landscape—rule-based expertise, experience-based problem solving, and data-driven learning. Quantum computing breakthroughs didn’t define AI in that era, networks without backpropagation reflect earlier work before the revival, and symbolic AI dominance doesn’t reflect the mixed, multi-paradigm reality of the time.

Three defining trends defined AI in the 1980s: expert systems, case-based reasoning, and the revival of neural networks with backpropagation. Expert systems represented domain knowledge as explicit rules and used inference engines to apply that knowledge, which led to practical, commercial AI applications in areas like diagnostics and configuration. Case-based reasoning offered a different approach by solving new problems through reusing and adapting solutions from similar past cases, providing flexible problem-solving for ill-defined tasks. At the same time, neural networks experienced a revival because the backpropagation algorithm made training multi-layer networks feasible, enabling learning from data and enabling patterns to be discovered without explicit programming of rules. These three threads together capture the decade’s distinctive AI landscape—rule-based expertise, experience-based problem solving, and data-driven learning. Quantum computing breakthroughs didn’t define AI in that era, networks without backpropagation reflect earlier work before the revival, and symbolic AI dominance doesn’t reflect the mixed, multi-paradigm reality of the time.