This course will introduce students to ways of thinking about how recent developments in AI systems powered by deep learning will shape everyday life and how to design such systems in manners that can respect human values. Students will read and discuss papers in Human-AI interaction, including but not limited to

1. Human-AI interactive systems that work / clash with the strengths and weaknesses of human cognition

2. Designing interactive, human-in-the-loop approaches in AI systems

3. Supporting interpretability, transparence, trust, and fairness in AI systems.

These topics will be explored in the context of real-world applications, through which students will learn how to think both optimistically and critically of what AI systems can do, and how they can and should be integrated into society.

This course will include a final project where students will form interdisciplinary groups to design an interactive Human-AI system powered by deep learning architectures of choice (vision, audio, language, etc.) most familiar to the group members. .

Prerequisite skills:

At the minimum, students need to have a basic knowledge of deep / machine learning, statistics, and an intermediate proficiency in python programming. This is NOT a machine learning or data mining course. Students who have previously taken advanced machine learning or deep learning courses should be prepared to apply what they have previously learnt in this course.