Generative Artificial Intelligence

3 units

This is a 3-credit introductory course to generative AI. The course assumes basic programming skills in Python, which can be acquired through the prerequisite data science course, DAT 561. The course is divided into four components: 1. Basic Technology in Natural Language Processing (NLP): This component covers the foundational technology in NLP that powers large language models. Topics include an introduction to PyTorch/Deep Learning, traditional NLP models (such as word vectors, transformer models, tokenizers, etc.), and the basic building blocks of foundational models. 2. Generative AI Usage and Programming: This component focuses on the practical applications of generative AI, enabling students to build business applications. Topics include an introduction to large language models (LLMs), prompt engineering, LLM evaluation, retrieval-argument generation (RAG), fine-tuning, AI agents and agentic workflows, and LLMOps. The goal is for students to reach a level where they can develop interesting business applications using these tools. 3. Societal and Business Impact of Large Language Models: This component addresses the broader impact of generative AI. It begins with an overview of the generative AI supply chain, from power generation and data center construction to end-user applications. The goal is to help students understand the short-term and long-term business impacts of generative AI. It also covers the societal impact and discusses how to build safe and responsible AI. 4. Beyond Large Language Models: This final component explores image, voice, and video generation. Topics include multimodal generative AI and diffusion models for image generation.