Jeff Clune Joins Vector Institute to Develop Open-Ended AI Systems
Jeff Clune, an associate professor at the University of British Columbia, has joined the Vector Institute for Artificial Intelligence as a new faculty member. His research focuses on creating open-ended AI systems capable of continuous learning and innovation, drawing inspiration from evolutionary biology and human culture. Unlike traditional machine learning models that optimize for single solutions, Clune pioneers quality diversity algorithms that generate diverse, high-quality behaviors. His notable contributions include a 2015 Nature paper on resilient robot locomotion and the development of Go-Explore, an algorithm designed to solve complex environments with sparse feedback. Additionally, he introduced POET, a system that creates and solves its own challenges in an open-ended stream. Clune’s ultimate ambition is to develop AI-generating algorithms that can endlessly improve themselves, potentially leading to human-level or super-human artificial intelligence. By marrying insights from natural evolution with advanced deep reinforcement learning techniques, Clune aims to replicate the endless inventive power of Darwinian evolution within computer systems, marking a significant step forward in the field of artificial intelligence research.
Wire timeline
Jeff Clune Joins Vector Institute to Develop Open-Ended AI Systems
Jeff Clune, an associate professor at the University of British Columbia, has joined the Vector Institute for Artificial Intelligence as a new faculty member. His research focuses on creating open-ended AI systems capable of continuous learning and innovation, drawing inspiration from evolutionary biology and human culture. Unlike traditional machine learning models that optimize for single solutions, Clune pioneers quality diversity algorithms that generate diverse, high-quality behaviors. His notable contributions include a 2015 Nature paper on resilient robot locomotion and the development of Go-Explore, an algorithm designed to solve complex environments with sparse feedback. Additionally, he introduced POET, a system that creates and solves its own challenges in an open-ended stream. Clune’s ultimate ambition is to develop AI-generating algorithms that can endlessly improve themselves, potentially leading to human-level or super-human artificial intelligence. By marrying insights from natural evolution with advanced deep reinforcement learning techniques, Clune aims to replicate the endless inventive power of Darwinian evolution within computer systems, marking a significant step forward in the field of artificial intelligence research.
Vector Institute for Artificial Intelligence