Foundations of Generative AI

This unit surveys the historical evolution of artificial intelligence and the core technical foundations of machine learning, deep learning, and transformer-based generative models. It introduces neural network architectures and AWS considerations for training and deploying deep learning models to prepare students for hands-on GenAI development in later units.

Learning Objectives

• Analyze the historical development of AI from mechanical reasoning through symbolic and subsymbolic eras to modern deep learning and LLMs, identifying key milestones and paradigm shifts.
• Compare and contrast symbolic AI, subsymbolic approaches, and deep learning by citing representative algorithms, strengths, limitations, and typical application domains.
• Demonstrate understanding of fundamental machine learning and deep learning concepts (including model components, training process, evaluation metrics, and common neural network architectures).
• Apply foundational knowledge to design a high-level AWS-based workflow for training and deploying a generative AI model, specifying appropriate services, compute/storage considerations, and evaluation criteria.