Topic for 2024-25: From Perceptrons to ChatGPT: How Computational Models Help Us Understand the Mind

Cognitive scientists have used mathematical and computational methods to understand human cognition since at least the 1940s. Similarly, the study of human neuroscience and cognition has influenced the development of artificial intelligence systems. Beginning in the early 2010s, massive increases in computational power and the accessibility of large databases have resulted in the rapid rise of human-like artificial intelligence systems, culminating in well-known public AI tools like ChatGPT. To what degree are these models a reflection of human intelligence, and can they help us understand human cognition? Are human-like cognitive biases also present in these models, and does this present ethical issues with their use? This course will cover the history of computational modeling in cognitive science, from early debates about modularity, interactivity, and the nature of representation; to the modern development of deep neural networks not only as practical systems, but as models of human cognition.

Units: 1

Max Enrollment: 15

Crosslisted Courses:

Prerequisites: Open to Juniors and Seniors who have taken one of the following - PSYC 215, CLSC 216/PSYC 216, PSYC 217, PSYC 218, PSYC 219, LING 114, PHIL 215, or CS 111; or permission of the instructor.

Instructor: Bushong

Distribution Requirements: EC - Epistemology and Cognition; SBA - Social and Behavioral Analysis

Typical Periods Offered: Spring

Semesters Offered this Academic Year: Spring

Notes: This is a topics course and can be taken more than once for credit as long as the topic is different each time.