The course will focus on understanding the nuances of interactive AI-based systems, their benefits for the final users, and the potential problems they can generate if they are naively designed. The emphasis will be on understanding how Human-Computer Interaction (HCI) principles and methods can help design “human-centered” AI systems, which implies considering who AI systems are built for, designing AI models that can intrinsically support transparency, and evaluating through proper methods how well the resulting systems work with respect to user needs and expectations. After introducing the basic concepts of HCI, the course will focus on models, methods, and interaction paradigms peculiar to AI systems. The course includes a tutorial on designing and prototyping interfaces for knowledge exploration using Large Language Models.
Target audience
Postgraduate students and professionals with previous coursework on AI (basics)
Prerequisites
Basic programming skills, Python, LLMs, Prompt Engineering
Intended learning outcomes
Creating Value for Humans through AI: Students will be able to demonstrate theoretical and methodological knowledge for designing advanced interactive systems based on AI models
Teaching and learning methods
Lectures on theoretical topics; Recommended readings; Practical sessions (tutorials, guidelines) dedicated to the design and prototyping of user interfaces for AI-based systems.
Assessment
Multiple-choice quizzes at the end of each session
Syllabus
Introduction
- Introduction to Human-AI Interaction.
- Basic principles of Human-Centered Design, Usability and User eXperience.
Designing Human-AI Interaction
- Artificial Intelligence vs Intelligence Augmentation
- Suggested readings on “substitution vs. augmentation”
- Paradigms for Human-AI Interaction: smart interface vs. conversational (with or without anthropomorphism and its effect: uncanny valley)
- Explainability for AI systems
- Suggested readings on “Explainability”
Hands-on session
- Tutorial on designing and prototyping interfaces for knowledge exploration using Large Language Models
Instructor
Maristella Matera
Politecnico di Milano
Hint Lab (Human-Centric Interactive Technologies Lab)
About Instructor
