Artificial Intelligence

🏠 ⮐

Artificial Intelligence (AI) refers to the capability of computational systems to perform tasks that typically require human intelligence. These tasks include reasoning, learning from experience, understanding natural language, recognizing patterns, solving problems, and making decisions. AI systems process information, identify correlations, and generate outputs based on algorithms and data rather than explicit programmed instructions for every scenario.

The field emerged formally in the 1950s with foundational work by pioneers like Alan Turing, who proposed the concept of machine intelligence, and John McCarthy, who coined the term "artificial intelligence" at the 1956 Dartmouth Conference. Since then, AI has evolved through multiple waves of development, characterized by periods of intense optimism ("AI summers") and reduced funding and interest ("AI winters").

Modern AI encompasses two broad categories: narrow AI (or weak AI), which excels at specific tasks like image recognition, language translation, or game playing, and general AI (or strong AI), a theoretical form that would possess human-like cognitive abilities across diverse domains. Current AI systems are exclusively narrow, highly specialized for particular applications.

AI systems operate through various approaches. Symbolic AI uses explicit rules and logic to manipulate symbols representing knowledge. Statistical and machine learning approaches identify patterns in data and improve performance through exposure to examples. Neural networks, inspired by biological brain structure, process information through interconnected layers of nodes. Recent advances in deep learning, which uses neural networks with many layers, have driven breakthroughs in computer vision, natural language processing, and generative systems.

Methodologies

AI methodologies are distinguished by their approach to knowledge and problem-solving: learning from data, encoding explicit rules and logic, evolutionary optimization, or planning through search and reasoning.

Symbolic AI approaches:

  • Knowledge Representation and Reasoning – Formalizing knowledge about the world in structured formats (logic, ontologies, rules) and using inference mechanisms to derive new conclusions or make decisions
  • Search and Optimization – Finding optimal or satisfactory solutions by systematically exploring problem spaces, including pathfinding, constraint satisfaction, and optimisation of objective functions
  • Evolutionary Computation – Optimisation algorithms inspired by biological evolution, using mechanisms like mutation, crossover, and selection to iteratively improve candidate solutions
  • Planning and Decision Making – Determining sequences of actions to achieve goals, reasoning about future states, and making choices under uncertainty or with incomplete information

Sub-symbolic/Statistical AI approaches:

  • Machine Learning – Systems that improve performance on tasks through experience and data, learning patterns and relationships without explicit programming for every scenario.

Hybrid approaches: Neuro-symbolic AI, some Evolutionary Computation methods