The Main Models of AI
Abbas Tcharkhtchi, Hamid Reza Vanaei, Sofiane Khelladi · 2025
This chapter presents a conceptual map of the major models that structure the field of AI. It begins by outlining the symbolic approach, where intelligence is modeled through logic, rules, and explicit representations. This paradigm dominated early AI research, particularly in expert systems and formal reasoning. The chapter then introduces the connectionist approach, which draws inspiration from NNs and emphasizes learning from data through adaptive weights and layers. It also examines the evolutionary paradigm, based on algorithms that simulate natural selection to optimize problem-solving strategies. Each model is analyzed in terms of its theoretical foundations, technical architecture, strengths, and limitations. The chapter further highlights hybrid approaches and distributed systems, including multi-agent models that combine symbolic reasoning with emergent behaviors. By comparing these frameworks, the chapter offers readers a clear understanding of the diversity within AI research and the philosophical questions underlying attempts to formalize intelligence in computational terms.