Abstract
This course offers an introductory yet comprehensive survey of the foundational principles, methods, and applications of Artificial Intelligence (AI) for first-year engineering students. It traces the historical evolution of AI and builds an understanding of core learning paradigms, including predictive and generative modeling, decision-making systems, and modern large-scale architectures. Emphasis is placed on conceptual understanding, practical relevance, and critical thinking around the technological and societal implications of AI. A final practical component introduces students to the use of contemporary large language models and retrieval-based AI systems.
Teaching and Learning Methods
Lectures and hands-on tutorials
Course Policies
Mandatory presence (additional details shared during the first lecture)
Bibliography
Russell, S. J., & Norvig, P. (2021). Artificial Intelligence: A Modern Approach (4th ed.)
Luger, G. F. & Stubblefield, W. (2009). Artificial Intelligence: Structures and Strategies for Complex Problem Solving
Prerequisites
Basic math (calculus, probability, linear algebra, ...)
Description
- Foundations of AI – History, Concepts, and Architectures. Introduces the origins and evolution of AI, the shift from symbolic to data-driven methods, and the conceptual foundations of learning and deep neural architectures.
- Learning to Predict – Supervised Learning Concepts. Explains how AI systems infer input-output mappings from data using predictive models, highlighting key principles like training, generalization, loss minimization, and model complexity.
- Machines That Generate – Images, Text, and Creativity. Covers generative modeling approaches that produce new data, contrasting autoregressive and latent-based methods, and exploring how models learn to synthesize coherent content.
- Decision-Making AI. Presents agents that learn to act through interaction, introducing reinforcement learning, exploration-exploitation trade-offs, and preference-based feedback mechanisms.
- Engineering AI Systems – Hardware, Frameworks, and Pipelines. Surveys the computational infrastructure behind AI, including GPUs, software libraries, and model deployment workflows critical for training and inference.
- Ethics and Safety in AI. Examines societal and technical challenges of AI deployment, addressing issues like bias, transparency, control, and long-term safety.
- Practical AI – Using Large Language Models and Retrieval-Augmented Generation. Offers hands-on exposure to modern AI tools, showing how LLMs interact with structured knowledge via prompts and retrieval mechanisms.
Learning Outcomes
- Describe the fundamental concepts of modern AI systems, including prediction, generation, and decision-making, and distinguish between the major learning paradigms that support them.
- Explain how AI models learn from data, including the roles of representation, supervision, evaluation, and system design in building effective learning systems.
- Identify and discuss key challenges in AI, including ethical risks, safety concerns, and the limitations of current technologies in real-world contexts.
Evaluation
Final exam (100%)