University of California, Berkeley | Fall 2026
Instructor: Serina Chang (serinac@berkeley.edu)
Time: TuTh 2:00-3:30pm
Location: Dwinelle 243
Office Hours: By appointment
UC Berkeley students: If you are interested in taking this course and can't directly enroll, please (1) join the waitlist if you can and (2) fill out this form. You're strongly encouraged to come to the first class (Thursday August 27) regardless of your enrollment status!
This course will explore the intersection of machine learning (ML) and human behavior. The format of the course will be a mix of paper presentations, lectures, and a final project. We will cover three units:
Unit I: Modeling Human Behavior with ML. Predicting or simulating human behaviors is useful when behaviors are difficult to observe (e.g., for cost or privacy reasons) or cannot be observed (e.g., future or counterfactual behaviors). In this unit, we'll begin with recent efforts to simulate behaviors with LLMs, including survey responses, experimental results, and social interactions. We will discuss challenges in this domain, such as bias, diversity, validation, generalization, and scalability, along with approaches to address these challenges. We will start from the individual-level and scale up to entire networks and societies, exploring both LLM-based and more traditional approaches to modeling human societies.
Unit II: Integrating Human Input into AI Development. In this unit, we will examine how human input is integrated into generative AI development, and how these choices shape downstream human-AI interaction. We will consider different points of integration, including post-training algorithms and inference-time context, along with limitations of existing approaches. We will then explore various ways in which AI needs to adapt to humans in order to improve human-AI outcomes, such as how to understand human intents, handle long-term interactions, learn individual preferences and personalize models, prioritize genuine human benefit over immediate satisfaction, and strive for human-AI complementarity.
Unit III: Algorithmically Infused Societies. In this final unit, we will expand from dyadic human-AI interaction to ML algorithms deployed at scale in society, and examine how these algorithms shape societal behaviors and outcomes. We will study these algorithms in different contexts, ranging from social media to search and recommender systems to modern generative AI. We will investigate feedback loops between algorithms and society—how algorithms shape human behavior while human behavior, in turn, feeds back into algorithms—and explore human awareness of algorithms. We will discuss societal impacts of algorithms along dimensions such as polarization, distribution of attention, and diversity in content creation and exposure.
Coming soon! See last year's course for an example of topics and readings.
Course topics were inspired in part by Johan Ugander's course on Social Algorithms and Joon Sung Park's course on AI Agents and Simulations. The course form was adapted from Sewon Min's course on Data-Centric Large Language Models.