Machine learning represents a paradigm shift from traditional programming, where algorithms learn patterns from data rather than following explicit instructions. This chapter examines the evolution from symbolic AI approaches to connectionist models, tracing the development from early perceptrons through contemporary deep learning systems. We explore three primary learning paradigms—supervised, unsupervised, and reinforcement learning—and their philosophical implications for understanding intelligence and categorization. The discussion includes the bias-variance tradeoff, feature representation, and the emergence of reinforcement learning with human feedback (RLHF) as a method for aligning AI systems with human preferences. Finally, we examine the transition from traditional machine learning to deep learning architectures that form the foundation of modern large language models.