Nick Haber | Motivation, Representation, and Autonomous Agents
The past decade’s series of dramatic AI successes has brought us closer to realizing the dream of the autonomous agent—an artificial system that can learn about its world, make decisions, and set and achieve goals, all with minimal human intervention.
Nature has given us an incredible motif for the autonomous agent. Humans are born with core motivations that drive behavior, and as they gain experience, they develop increasingly sophisticated representations of their worlds, which in turn facilitate increasingly complex behaviors.
How should we attempt to realize this motif for intelligence to create autonomous agents? In this seminar, the speakers will cover a number of efforts aimed at assembling pieces of it—starting with their own research into intrinsic motivation and adaptive “world model” representations, including work inspired by biology and, in particular, human development. They will also describe some of the opportunities language models provide for us to make increasingly autonomous systems.
Related Events
Dan Iancu & Antonio Skillicorn | Interpretable Machine Learning and Mixed Datasets for Predicting Child Labor in Ghana’s Cocoa Sector
Child labor remains prevalent in Ghana’s cocoa sector and is associated with adverse educational and health outcomes for children.
Child labor remains prevalent in Ghana’s cocoa sector and is associated with adverse educational and health outcomes for children.
Wolfgang Lehrach | Code World Models for General Game Playing
While Large Language Models (LLMs) show promise in many domains, relying on them for direct policy generation in games often results in illegal moves and poor strategic play.
While Large Language Models (LLMs) show promise in many domains, relying on them for direct policy generation in games often results in illegal moves and poor strategic play.