The Philosophy of AI

Catalog description

This course examines philosophical issues raised by artificial intelligence, with particular attention to recent advances in deep learning and large language models. It begins with the origins of modern deep learning in early neural network models and the connectionist approach, considering its premises, promises, and limitations. Explanations for the Deep Learning Revolution will then be considered. Subsequent topics include how to understand AI models as models of cognition, the interpretability of neural network models, whether language models have genuine mental states, and how to evaluate claims about alignment, agency, and consciousness. Students will engage with both technical and philosophical texts while developing their own analyses. No prerequisites, though some background in philosophy, computer science, or cognitive science is recommended.