Augustus is a PhD student in Electrical and Computer Systems Engineering, and his research aims to explore cognition from the side of Artificial Intelligence. He is working on model-based reinforcement learning; i.e. the creation of agents that learn to understand the dynamics of their environment, where these agents receive only limited feedback about their performance. He hopes that this will provide insights into how cognitive systems can come to represent environments, especially semantically meaningful properties like physics, affordances and causation. A deeper understanding of “mental” representation is of interest to philosophy in that it might shed light on how rich conceptions of the world can be created and exploited, and it is of instrumental importance to AI researchers since such conceptions seem to be at the core of how many animals can generalise information to new contexts in sophisticated and abstract ways.
Primary PhD Supervisor: Tom Drummond
Associate Supervisor: Jakob Hohwy