This graduate-level seminar course will cover foundational works in artificial social intelligence, an emerging multi-disciplinary research field with the vision of advancing AI that can perceive human social signals, interpret and engage in interactions with humans, and navigate social norms with common sense. The course will offer an in-depth examination of both seminal research and recent advances that address fundamental technical challenges in modeling human social behavior, understanding multimodal human communication, and building socially-intelligent agents. This topic is inherently interdisciplinary, and the readings and discussion in this course will integrate perspectives from machine learning, natural language processing, computer vision, multimodal interaction, robotics, cognitive science, neuroscience, and sociology. The course emphasizes critical reading, discussion, and engagement with the literature and provides opportunities for students to pursue semester research projects. Students are expected to have prior experience in artificial intelligence and a strong interest in the science of social intelligence.

  • Time: Monday/Wednesday 9:30-10:50am
  • Location: DH 1117
  • Assignment Submission: CMU Canvas
  • Contact: Please contact the instructors if you have any questions before the course Canvas launches. Include [11-866] in your email subject line.

Previous offerings: This course was taught in Spring 2023. The Spring 2026 version of 11-866 is being reintroduced with updated content, in particular on the computational side regarding social intelligence modeling paradigms.

For LTI Students: This course has been approved to fulfill the LTI Application Engineering requirement and the Human Language and Communication Breadth Requirement (12 units). For additional information, please contact your program director.